Sequestration efficiency in the iron-limited North Atlantic: implications for iron supply mode to fertilized blooms

Seques a ion e ficiency in he i on-limi ed No h A lan ic: Implica ions o i on supply mode o e ilized blooms F édé ic A. C. Le Moigne 1 , C. Ma k Moo e 2 , Richa d J. Sande s 1 , Ma ia Villa-Al ageme 3 , Sebas ian S eigenbe ge 2 , and E ic P. Ach e be g 2,4 1 Na ional Oceanog aphy Cen e, Sou hamp on, UK, 2 Uni e i y o Sou hamp on, Sou hamp on, UK, 3 Depa men o Físíca Aplicada II, Uni e sidad de Se illa, Spain, 4 GEOMAR, Helmhol z Cen e o Ocean Resea ch Kiel, Ge many Abs ac Es ima es o he amoun o ca bon seques e ed in he ocean in e io pe uni i on (Fe) supplied, as quan ified by he seques a ion e ficiency (C e x ), a y widely. Such a iabili y in C e x has equen ly been a ibu ed o es ima e unce ain y a he han in insic a iabili y. He e we de i e new es ima es o C e x o he subpola No h A lan ic, whe e Fe s essed condi ions ha e ecen ly been demons a ed. De i ed alues o C e x om ac oss he egion, including a eas subjec o a ypical ex e nal Fe e iliza ion e en s du ing he yea o sample collec ion (2010), anged om 17 o 19 kmol C (mol Fe 1 ). Compa ing hese es ima es wi h alues om o he sys ems, conside ed in he con ex o a iable bloom du a ions in he di e en oceanog aphic se ings, we sugges ha appa en a iabili y in C e x may be ela ed o he mode o Fe deli e y. 1. In oduc ion I on (Fe) a ailabili y has been shown o con ol phy oplank on g ow h in he so-called high ni a e low chlo ophyll (HNLC) egions [Blain e al., 2007; Boyd e al., 2004; de Baa e al., 2005; Polla d e al., 2009; Sme acek e al., 2012] including he Sou he n Ocean, equa o ial Pacific, and subpola No h Pacific. Mo e ecen ly, some oceanic egions cha ac e ized by ma ked seasonal chlo ophyll peaks (blooms), including he I minge and Iceland Basins (he ea e IRB and IB, espec i ely) o he high la i ude No h A lan ic (HLNA), [Nielsdo i e al., 2009; Ryan-Keogh e al., 2013; Sande s e al., 2005] ha e also been sugges ed o expe ience a deg ee o Fe s ess. Wi hin all hese egions, Fe deficiency po en ially con ibu es o incomple e u iliza ion o su ace mac onu ien s. Subsequen subduc ion o hese unused mac onu ien s in egions o deep wa e o ma ion, including he Sou he n Ocean o he HLNA, ep esen s an ine ficiency in he biological ca bon pump [Ma ino e al., 2008; Nielsdo i e al., 2009; Sa mien o and O , 1991]. Consequen ly, oceanic Fe a ailabili y has been in oked as a po en ial con ol on a mosphe ic CO 2 [Ma in e al., 1990], wi h o example, highe ae osol deposi ion o he glacial Sou he n Ocean hypo hesized o ha e pa ly con ibu ed o glacial/in e glacial CO 2 cycles [Jickells e al., 2005; Ma inez-Ga cia e al., 2014; Ridgwell and Wa son, 2002]. Quan i a i e unde s anding o linkages be ween a iabili y in ex e nal Fe inpu s and ca bon cycling in di e en oceanic se ings equi es an unde s anding o he ela ionship be ween Fe supply and ca bon seques a ion, as encapsula ed in he seques a ion e ficiency (C e x ). To da e, es ima es o C e x (defined as he a io o ca bon expo ed pe uni o Fe supplied) om field p og ams ha e a ied widely (~1.2 o 154 kmol C (mol Fe) 1 )[Boyd e al., 2007; de Baa e al., 2005; Mo is and Cha e e, 2013]. The mode o Fe supply has p e iously been sugges ed as a mechanis ic d i e o such a iabili y in C e x [Boyd e al., 2007; Che e e al., 2010]. Howe e , in he absence o any obse ed sys ema ic basis o he epo ed >2 o de o magni ude ange in C e x , hese di e ences may also eflec unce ain ies in calcula ions [Mo is and Cha e e, 2013], including incomple e accoun ing o Fe sou ces o di e ing app oaches o es ima ing C expo . Wi h he excep ion o alues de i ed o Fe e ilized blooms a ound (sub-) An a c ic island sys ems, such as epo ed o he KE guelen Ocean and Pla eau compa ed S udy (KEOPS) and C oze na u al i on bloom and expo expe imen (CROZEX) s udies [Blain e al., 2007; Polla d e al., 2009], no es ima es o C e x ha e ye been de i ed o highly p oduc i e na u al oceanic sys ems. He e we epo new es ima es o C e x o he HLNA and a emp o syn hesize hese in o a g owing unde s anding o po en ial con ols on his a iable in he global ocean. LE MOIGNE ET AL. ©2014. The Au ho s. 4619 PUBLICATION S Geophysical Resea ch Le e s RESEARCH LETTER 10.1002/2014GL060308 Key Poin s: •Seques a ion e ficiency om No h A lan ic •Seques a ion e ficiency may be ela ed o he mode Fe supply •Mode o Fe supply needs o be conside ed in models Suppo ing In o ma ion: •Readme •Le_Moigne_e _al_GRL_sup_June12. docx Co espondence o: F. A. C. Le Moigne, [email p o ec ed] Ci a ion: LeMoigne,F.A.C.,C.M.Moo e, R. J. Sande s, M. Villa-Al ageme, S. S eigenbe ge , and E. P. Ach e be g (2014), Seques a ion e ficiency in he i on-limi ed No h A lan ic: Implica ions o i on supply mode o e ilized blooms, Geophys. Res. Le .,41, 4619–4627, doi:10.1002/2014GL060308. Recei ed 23 APR 2014 Accep ed 18 JUN 2014 Accep ed a icle online 20 JUN 2014 Published online 8 JUL 2014 This is an open access a icle unde he e ms o he C ea i e Commons A ibu ion License, which pe mi s use, dis ibu ion and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed. 2. Me hods o Ancilla y Da a and Expo Fluxes Sampling ook place om 4 July o 10 Augus 2010, on boa d he RRS Disco e y c uise D354 as pa o he I minge Basin I on S udy (IBIS) p og am. Ino ganic nu ien s and pa icula e o ganic ca bon (POC) and ni ogen (PON) we e sampled and analyzed as p e iously desc ibed [Le Moigne e al., 2013b; Sande s and Jickells, 2000]. POC and PON expo fluxes we e subsequen ly calcula ed using he 234 Th “small- olume” echnique [Pike e al.,2005].B iefly, e ical p ofiles o 234 Th ac i i y (in eg a ed o a dep h o 100 m) we e con e ed o es ima es o downwa d 234 Th flux using a one dimensional s eady-s a e model [Buessele e al., 1992]. Ex ac ion e ficiencies o 234 Th we e 90.6 ± 6.7% [Le Moigne e al., 2013a, 2012]. Obse ed 234 Th:POC and 234 Th:PON a ios o la ge (>53 μm) pa icles collec ed using in si u S and Alone Pumping Sys ems (SAPS) deployed o 1.5 h a a single dep h benea h he mixed laye we e hen used o con e 234 Th fluxes o POC/N fluxes. App oxima ely 1500–2000 L o seawa e was fil e ed using 53 μmmeshfil e s (Ni ex), wi h swimme s manually emo ed ollowing fil a ion. Pa icles we e hen insed o he fil e s using Th- ee seawa e as p epa ed ollowing [Le Moigne e al., 2013b], and he pa icle suspensions we e spli andanalyzed o 234 Th, POC, and PON as desc ibed in Le Moigne e al. [2013b]. Mai i e al. [2012] sugges ed ha high flow a es may lead o pa icle disin eg a ion. The pump a es we used we e abou wo imes la ge han hose used by Mai i e al. [2012];howe e , hesu acea eao ou fil e s was a ound 4 imes la ge [Le Moigne e al., 2013b]. The e o e, he eloci y o seawa e h ough he fil e was a ound hal o ha used by Mai i e al. [2012]. Dissol ed i on was de e mined ollowing p ocedu es epo ed by Pain e e al. [2014]. 3. Resul s and Discussion 3.1. The High La i ude No h A lan ic in 2010 Biogeochemical cycles in he HLNA we e pe u bed du ing 2010 as a esul o ae osol Fe deposi ion o he IB om he e up ion o he Icelandic olcano Eyja jallajökull [Ach e be g e al., 2013]. In addi ion, win e mixed laye dep hs p io o ou c uise we e shallowe han a e age in he IRB, while he summe eupho ic zone (40 ± 5 m) and mixed laye dep hs (28 ± 8 m) ep esen ed ypical condi ions o bo h he IB and IRB [Henson e al., 2013; Pain e e al., 2014]. Consequen ly, inpu s o Fe om deep win e con ec ion we e also likely up o ou old highe in he IB (37,500 nmol m 2 ) ela i e o he IRB (10,000 nmol m 2 )[Pain e e al., 2014]. Volcanic Fe inpu s du ing May 2010 o he IB po en ially esul ed in enhanced mac o-nu ien d awdown, wi h low obse ed ni a e(<1μM) du ing summe 2010 in he IB [Ach e be g e al.,2013;Ryan-Keogh e al., 2013], while concen a ions emained ela i ely high (3–5μM) and compa able wi h p e ious obse a ions in he IRB [Sande s e al., 2005] (Figu es 1a and 1c). Silica e was also s ongly deple ed in he IB (<1μM) ela i e o IRB (1–4μM) (Figu es 1b and 1d). Co espondingly, nu ien en ichmen expe imen s demons a ed clea e idence o Fe limi a ion in he IRB du ing summe , while Fe s ess was much less se e e unde he low mac onu ien condi ions encoun e ed in he IB [Ryan-Keogh e al., 2013]. Fu he mo e, phy oplank on Fe:C up ake a ios calcula ed using adio ace inco po a ion echniques [Poul on e al., 2010; Twining e al., 2004] a e aged 0.9 (±0.6) and 4.1 (±1) μmol mol 1 du ing summe in he IRB and IB, espec i ely (C.M.Moo eandA.J.Poul on,unpublishedda a),compa able wi h obse a ions o Fe-limi ed empe a e axa (~2–10 μmol mol 1 )[Sunda and Hun sman, 1995] and, o he IRB, Fe-limi ed axa isola ed om ch onically low Fe en i onmen s [S zepek e al., 2012]. By he ime o ou summe sampling, chlo ophyll-a (Chl-a) concen a ions in he IB had dec eased ollowing he sp ing bloom [Ryan-Keogh e al., 2013]. In con as , a ma ked bloom (~1–4μgl 1 ) was s ill unde way in he cen al IRB, po en ially as a esul o anomalous hyd og aphic o cing [Henson e al., 2013]. Condi ions in he pos -bloom wes e n IRB [Ryan-Keogh e al., 2013] we e simila o hose wi hin he classical Fe-limi ed sys ems, wi h ela i ely low Chl-a concen a ions (<1μgl 1 ) and esidual ni a e (>4μM). He ea e , we hus es ic discussion o he IRB o his wes e n egion [Ryan-Keogh e al., 2013]. Consequen ly, we compa e ca bon expo measu emen s o wo s a ions in he IRB (s a ions 10 and 16, he ea e lowe Fe inpu , Fe) and h ee s a ions in he IB (s a ions 6, 28, and 33, he ea e highe Fe inpu , +Fe) (Figu e 1). 3.2. Ca bon Expo In o de o de i e C e x , es ima es o bo h ca bon expo and Fe inpu fluxes a e equi ed. Following ou p e ious wo k [Polla d e al., 2009], seasonal nu ien defici s we e combined wi h es ima es o he 234 Th- based C and N Geophysical Resea ch Le e s 10.1002/2014GL060308 LE MOIGNE ET AL. ©2014. The Au ho s. 4620 19448007, 2014, 13, Downloaded om h ps://agupubs.onlinelib a y.wiley.com/doi/10.1002/2014GL060308 by Spanish Coch ane Na ional P o ision (Minis e io de Sanidad), Wiley Online Lib a y on [28/11/2024]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License expo fluxes [Buessele e al., 1992; Le Moigne e al., 2013b]. Di iding he obse ed ino ganic N defici (Figu e 2a) wi h ho ium-de i ed es ima es o downwa d N flux, hen mul iplying he esul an es ima es o expo du a ion by he ho ium-de i ed daily o ganic ca bon fluxes (Figu e 2b) (see suppo ing in o ma ion), we de e mined annually in eg a ed POC expo es ima es o 780 and 1330 mmol m 2 in he IRB (Fe) and he IB (+Fe), espec i ely (Figu e 2c, ange o bo h +Fe and Fe gi en in Table 1). Excess C expo be ween he Fe- eple e (IB) and Fe-limi ed (IRB) a ea (Figu e 2c) was hus 550 mmolm 2 , simila o a p e ious es ima e o he C oze egion [Mo is and Cha e e,2013;Polla d e al., 2009], bu lowe han ha o he KEOPS expe imen [Blain e al., 2007]. 3.3. Sou ces o New I on In keeping wi h p io wo k [Blain e al., 2007; Polla d e al., 2009], we conside new sou ces o Fe o he sys em when de i ing C e x, while acknowledging ha ecycled Fe was likely impo an in suppo ing addi ional ecycled p oduc ion [Boyd and Ellwood, 2010]. As summa ized in Table 1, we conside ed fi e po en ial sou ces o new Fe o he IB and IRB du ing 2010 (see suppo ing in o ma ion o de ailed de i a ions), namely: (1) ypical (i.e., non olcanic) a mosphe ic deposi ion [Ach e be g e al., 2013], (2) addi ional olcanic a mosphe ic Fe flux o he IB as a esul o he Eyja jallajökull e up ion [Ach e be g e al., 2013; Pain e e al., 2014], (3) upwa d e ical di usi e flux [Pain e e al., 2014], (4) con ec i e win e mixing flux [Pain e e al., 2014], and (5) ho izon al Fe flux om he adjacen con inen al shel es. The la e may include ma e ial o igina ing om ben hic supplies om con inen al shel es [El od e al., 2004] and di ec uno om i e s and glacial mel [Bha ia e al., 2013; Hawkings e al., 2014]. Ho izon al fluxes we e es ima ed by conside ing su ace wa e Fe concen a ions away om shel egions, idges, o majo cu en s in bo h he IB and he IRB ollowing p e iously app oaches [Buccia elli e al., 2001; Planque e e al., 2007; Rijkenbe g e al., 2012] (see suppo ing in o ma ion), conside ing wo main sou ces o he IB (Icelandic shel and Reykjanes idge) and h ee main sou ces o he IRB (G eenland shel , he Reykjanes idge, and he Eas e n G eenland Cu en ). O e all, ele a ed dissol ed and pa icula e i on (DFe and PFe) concen a ions (Figu es S2 and S3) obse ed close o he Iceland and G eenland shel es did no pe sis in o Figu e 1. Nu ien s in he high-la i ude No h A lan ic. (a) Map o su ace wa e ni a e+ni i e concen a ions (μM) du ing summe 2010. Ca bon expo s a ions a e indica ed o e e ence, 10 and 16 a e he high ni a e s a ions while 06, 28, and 33 a e he low ni a e s a ions. (b) Map o su ace wa e silica e concen a ions (μM) du ing summe 2010. (c) Ni a e + ni i e concen a ion (μM) p ofiles a high (blue filled, iangles: s 10; ci cles: s 16) and low (g een filled, ci cles: s 6; iangles: s 28, squa es: s 33) nu ien s s a ions. (d) Silica e concen a ion (μM) p ofiles a , s 6, 10, 16, 28, and 33, symbols a e in Figu e 1c (see Table S1, S2, and ex in suppo ing in o ma ion o ull desc ip ion). Geophysical Resea ch Le e s 10.1002/2014GL060308 LE MOIGNE ET AL. ©2014. The Au ho s. 4621 19448007, 2014, 13, Downloaded om h ps://agupubs.onlinelib a y.wiley.com/doi/10.1002/2014GL060308 by Spanish Coch ane Na ional P o ision (Minis e io de Sanidad), Wiley Online Lib a y on [28/11/2024]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License he cen al basins and hence ho izon al inpu s we e mino , albei s ill somewha unce ain (Table 1). Ho izon al fluxes and simila mino e ical di usi e fluxes we e scaled up by mul iplying daily fluxes by 100 days ( he es ima ed bloom du a ion), and his showed ha bo h we e ul ima ely mino con ibu o s o o e all annual new Fe inpu s (Table 1), which we e domina ed by he win e con ec i e supply [Nielsdo i e al.,2009;Pain e e al., 2014]. In o al, he annual inpu o new Fe o he IB (0.041–0.044 mmol m 2 ) was es ima ed o be ~4 imes la ge han he inpu o he IRB (0.011 mmol m 2 ), p incipally due o highe con ec i e and olcanic fluxes (Table 1). Bo h hese Fe inpu es ima es a e compa able wi h epo ed alues om he KEOPS (0.001–0.023 mmol m 2 ) and CROZEX (0.016–0.076 mmol m 2 ) s udies [Mo is and Cha e e, 2013] (Figu e 3a). 3.4. Seques a ion E ficiency Values o C e in he IB and IRB we e ~30 and ~65 kmol C (mol Fe) 1 , espec i ely (Table 2), u he confi ming ha he IB expe ienced Fe e iliza ion ( alues o C e we e calcula ed sepa a ely o he wo basins a he han as he di e ence be ween he wo basins, see suppo ing in o ma ion). Consis en ly wi h p e ious s udies [Blain e al., 2007; Polla d e al., 2009], we also calcula ed he appa en e ficiency o he excess expo esul ing om he enhanced flux o new Fe o he IRB (C e x ) (suppo ing in o ma ion), yielding alues o 17–19 kmol C (mol Fe) 1 , close o he CROZEX (17 kmol mol 1 ) bu lowe han KEOPS (154 kmol mol 1 ) alues (Table 2, no e he ea e we use he co ec ed alues [Che e e al., 2010; Mo is and Cha e e, 2013] o KEOPS which include an addi ional sou ce o Fe no conside ed in he seminal s udy [Blain e al., 2007]). All hese es ima es a e conside ably highe hose om h ee a ificial Fe e iliza ions expe imen s (1.2–6.5 kmol mol 1 , loca ions, and e e ences gi en in Table 2). P e ious s udies discussing such anges in es ima es o C e (x) ha e ended o ocus on he compa abili y and alidi y o he me hods used o cons uc Fe and C budge s [Blain e al., 2007; Che e e al., 2010; Mo is and Cha e e, 2013; Polla d e al., 2009], p esumably based on he inhe en assump ion ha he a iabili y in C e (x) was oo la ge o be asc ibed o na u al ac o s. 3.5. Mode o Fe Supply He e we ocus on an al e na i e hypo hesis, namely ha , as p e iously sugges ed [Boyd e al., 2007; Che e e al., 2010], he obse ed ange may ins ead be d i en by di e ences ela ed o he mode o Fe supply. Th ee dis inc modes o Fe supply cha ac e ize he a ious expe imen s. (1) In he shallow wa e s o he Ke guelen pla eau (500 m deep [Blain e al., 2007]) C e x was la ge a in e media e le els o Fe inpu s du ing Figu e 2. Annual ca bon expo in he Iceland and I minge basins. (a) POC and PON expo fluxes in eg a ed a 100 m (mmol m 2 d 1 ), (b) ni ogen budge (ni ogen defici mmol m 2 , and bloom du a ion, days, see main ex o explana ion), and (c) annual ca bon expo (mmol m 2 ). E o s (s anda d de ia ion) a e also gi en in Table 1. Geophysical Resea ch Le e s 10.1002/2014GL060308 LE MOIGNE ET AL. ©2014. The Au ho s. 4622 19448007, 2014, 13, Downloaded om h ps://agupubs.onlinelib a y.wiley.com/doi/10.1002/2014GL060308 by Spanish Coch ane Na ional P o ision (Minis e io de Sanidad), Wiley Online Lib a y on [28/11/2024]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License he KEOPS s udy, wi h con inued supply Fe h oughou he yea h ough a combina ion o diapycnal mixing ac oss a e ical g adien o e he shallow shel and a no hwes wa d flow om Hea d islands pas he eas e n side o he Ke guelen pla eau [Che e e al., 2010]. Hence, he Fe s ocks abo e he pla eau did no appea o a y much o e he du a ion o he bloom [Blain e al., 2008; Che e e al., 2010]. (2) Con e sely, in he deep wa e s o he p oduc i e HLNA and he e ilized egion o he CROZEX bloom, he main supply o Table 1. Fe Inpu s and Ca bon Budge in Bo h Iceland and I minge Basins (IB and IRB) Annual Flux o I on (nmol m 2 ) IRB (Fe) IB (+Fe) Lowe Limi b IB (+Fe) Uppe Limi b Re e ence Ca bon Budge IRB (Fe) IB (+Fe) A mosphe ic (1168) 3478 6128 [Ach e be g e al., 2013; Pain e e al., 2014] Ni a e d awdown (mmol m 2 ) 201 ± 25 287 ± 9 Ve ical di usi e flux 160 a 250 a 250 a [Pain e e al., 2014] Daily POC ex (mmol m 2 d 1 ) 7.0 ± 5.7 10.2 ± 4.1 Win e mixing 10,000 37,500 37,500 [Pain e e al., 2014] Daily PON ex (mmol m 2 d 1 ) 0.8 ± 0.6 1.5 ± 0.6 Ho izon al Fe 15 a 4.5 a 4.5 a Bloom du a ion (days) 111 130 To al 11,343 41,232 43,882 Annual POCex (mmol m 2 ) 778 1331 To al (in mmol m 2 ) 0.011 0.041 0.044 Range c (lowe and uppe limi s) 682–769 1266–1449 a This was mul iplied by 100 days o eflec he du a ion o he bloom (see ex in suppo ing in o ma ion). Values in b acke s a e measu ed “backg ound”fluxes [Pain e e al., 2014] (see ex in suppo ing in o ma ion). b The lowe and uppe limi o Fe flux o he IB eflec he sensi i i y analysis on he modeled olcano i on deposi ion gi en in Ach e be g e al. [2013]. c Calcula ed as [Polla d e al., 2009]. Figu e 3. Seques a ion e ficiency, annual ca bon expo , and nu ien s d awdown o six a ificial and na u al Fe e iliza- ion on he ocean. (a) Annual ca bon expo (mmol m 2 ) e sus annual Fe inpu (mmol m 2 ), e ilized egions a e in g een ci cles, and non- e ilized egions a e in blue ci cles. Expe imen names and e ilized/non e ilized egions a e indica ed (I: IBIS; K: KEOPS; C: CROZEX). (b) Ni a e d awdown (mmol m 2 ) (inpa ch ou pa ch concen a ions o ni a e mul iplied by expo in eg a ion dep h o SERIES, EIFEX, and SOFEX) e sus C e x and C e (mol mol 1 ); ed a e na u ally e ilized egions (ci cles: C e x , iangles: C e + Fe egions, squa es: C e Fe egions), and blue ci cles a e a ifical e iliza ions (C e x ) (see Table 2). The mode o Fe supply is indica ed on op o he panel as well as he Fe:C a io in phy oplank on ( e e ence in he main ex ). (c) Es ima ed expo du a ion (days, gi en in [Polla d e al., 2009] o CROZEX and in Figu e 2b o IBIS. KEOPS expo du a ion was es ima ed in a simila ashion by di iding he daily downwa d fluxes o POC by he “seasonaly in eg a ed”downwa d fluxes o POC gi en in Blain e al. [2007]] ( hei Table 1). Bloom du a ions o KEOPS a e 198 and 167 days, espec i ely, o + and Fe egions) e sus C e (mol mol 1 ); ed ci cles a e na u al e ilized/non- e ilized egions, and blue ci cles a e a ificial e iliza ions (see Table 2) concen a ions (in μM, es ima ed om nu ien s e e ences gi en in Table 2 and Le i us [1982]) and annual ca bon expo (mmol m 2 ). (d) Bloom du a ion in he +Fe egion (days) es ima ed om obse ed sa elli e-de i ed Chl-a ime se ies (KEOPS [Blain e al., 2007]; CROZEX [Polla d e al., 2009]; IBIS [Ach e be g e al., 2013], see Table S3 in suppo ing in o ma ion) e sus C e x . Bloom du a ion o a ificial expe imen s is assumed o be equal o he leng h o each expe imen [Boyd e al., 2004; Buessele e al., 2004; Sme acek e al., 2012]. Geophysical Resea ch Le e s 10.1002/2014GL060308 LE MOIGNE ET AL. ©2014. The Au ho s. 4623 19448007, 2014, 13, Downloaded om h ps://agupubs.onlinelib a y.wiley.com/doi/10.1002/2014GL060308 by Spanish Coch ane Na ional P o ision (Minis e io de Sanidad), Wiley Online Lib a y on [28/11/2024]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License Fe appea s o be seasonal [Polla d e al., 2009] (Table 1). Hence, in bo h hese egions, Fe migh be expec ed o build up wi hin he mixed laye du ing win e (a pe iod o abou 100 days). Once his mixed laye shoals in sp ing, he Fe pool can be used du ing he p oduc i e season un il exhaus ion [Nielsdo i e al., 2009; Planque e e al., 2007]. A simila si ua ion is likely o cha ac e ize much o he Sou he n Ocean [Boyd e al., 2012; Tagliabue e al., 2014]. (3) Finally, du ing he a ificial expe imen s, se e al ons o FeSO 4 we e supplied o he su ace ocean in one o se e al disc e e seedings o e a pe iod o se e al weeks, esul ing in a la ge loss o Fe due o he o ma ion o insoluble Fe oxy-hyd oxides [Boyd e al., 2000]. 3.6. Implica ion o he Mode o I on Supply in Fe ilized Blooms We sugges ha i is likely ha high C e x alues a e associa ed wi h p o ac ed blooms whe e Fe a ailabili y is main ained due o con inuous Fe supply and pe haps also Fe complexa ion by o ganic ligands, sus aining con inued C expo o e a p olonged pe iod, a he han being apidly los h ough p ecipi a ion and sca enging. We es he associa ion o high C e x wi h p olonged blooms by compa ing Fe supply modes o in si u es ima es o ni a e up ake, du a ion o he expo phase, and bloom du a ion (de i ed om sa elli e Chl-a ime se ies, see suppo ing in o ma ion Table S3). We acknowledge ha es ima es o he du a ion o expo o he HLNA and CROZEX a e no independen o he calcula ed seasonally in eg a ed C expo (and hus C e (x) ), as bo h a e based on obse ed ni a e d awdown. Howe e , ni a e d awdown and C e x we e independen a iables in KEOPS and in he a ificial Fe elease expe imen s. Mo eo e , bloom du a ions de i ed om sa elli e Chl-a ime se ies a e independen o C e x calcula ions in all cases and scaled well wi h es ima ed expo du a ions (see Table S3 in suppo ing in o ma ion). O e all, bo h C e x and/o C e , we e co ela ed wi h bloom (o expe imen ) ni a e up ake (Figu e 3b), expo du a ion (Figu e 3c), and bloom du a ion (Figu e 3d). Mo eo e , he maximum es ima es o C e x and/o C e we e compa able o he ange o C:Fe a ios measu ed in Fe- eple e phy oplank on cul u es [Sunda and Hun sman, 1995], which could concei ably se he uppe bound on C e x . All he ma kedly lowe es ima es o C e o da e hus appea o ela e o si ua ions whe e Fe supply was discon inuous (Figu e 3). Blooms cha ac e ized by a mo e con inuous supply o Fe (KEOPS) hus appea o expo ca bon mo e e ficien ly pe uni o Fe added o e a comple e seasonal cycle. Such an e ec would be consis en wi h a low bu con inuous supply o Fe allowing e ficien e en ion o Fe wi hin he dissol ed phase in he eupho ic zone, likely h ough ull complexa ion by o ganic ligands [Gledhill and Buck, 2012], o wi hin he ecosys em i sel . Indeed, biological up ake wi hin such blooms could be en isaged o gene a e he g adien s in bioa ailable Fe a he pe iphe y (ei he below o o he edges o he bloom), which would likely dic a e he magni ude o Fe flux in o he bloom egion. Such a scena io is in ma ked con as o he si ua ion in Table 2. Ca bon Seques a ion E ficiency and Nu ien D awdown A ea Type o Fe iliza ion Ni a e D awdown (mmol m 2 ) d Seques a ion E ficiency (Ce x and Ce , kmol mol 1 ) Re e ence C e x HLNA Na u al 287 17.2–19.0 This s udy Sou he n Ocean Na u al 687 b 154.0 a [Blain e al., 2007] Sou he n Ocean Na u al 404 c 17.2 a [Polla d e al., 2009] Sou he n Ocean A ificial 75 e 6.5 [Sme acek e al., 2012] Sou he n Ocean A ificial 75 e 3.3 [Buessele e al., 2004] Suba c ic PacificA ificial 200 e 1.2 [Boyd e al., 2004] C e HLNA +Fe (IB) gi en abo e 30.2–32.5 This s udy Fe (IRB) 201 65.5 Sou he n Ocean (KEOPS Ke guelen) +Fe gi en abo e 227.3 a [Blain e al., 2007] Fe 332 b 2,883.3 a Sou he n Ocean (CROZEX C oze ) +Fe gi en abo e 18/8 a [Polla d e al., 2009] Fe 161 c 25.0 a a Recalcula ed in [Mo is and Cha e e, 2013]. b Nu ien da a om KEOPS a e p esen ed in Mosse i e al. [2008]. c Nu ien da a om CROZEX a e p esen ed in Sande s e al. [2007] ( e e ence in suppo ing in o ma ion). d Ni a e d awdown in he uppe panel o he able is gi en o he e ilized a ea o pa ch (+Fe). e Nu ien d awdown was calcula ed as su ace nu ien in-pa ch su ace nu ien ou -pa ch mul iplied by expo in eg a ion dep h. Geophysical Resea ch Le e s 10.1002/2014GL060308 LE MOIGNE ET AL. ©2014. The Au ho s. 4624 19448007, 2014, 13, Downloaded om h ps://agupubs.onlinelib a y.wiley.com/doi/10.1002/2014GL060308 by Spanish Coch ane Na ional P o ision (Minis e io de Sanidad), Wiley Online Lib a y on [28/11/2024]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License pu pose ul FeSO 4 elease expe imen s, whe e a la ge amoun o Fe is supplied o a egion wi h low p e- exis ing biomass. O de s o magni ude lowe C e x in such si ua ions migh be expec ed o esul om loss o much o he pulsed excess Fe inpu s [Boyd e al., 2004; Buessele e al., 2004; Sme acek e al., 2012] due o sa u a ion o he a ailable Fe complexing ligands si es and apid ans e o Fe in o colloidal (>200 kDa) phases wi h subsequen pa icle sca enging alongside a limi ed capaci y o he ex an mic obial communi y o ake up he sudden la ge Fe excess [Bowie e al., 2001; Boyd e al., 2007]. Finally, in sys ems whe e he supply o Fe is seasonally domina ed (CROZEX, IBIS), he bloom likely e mina es when he Fe s ock is used up [Boyd e al., 2012; Nielsdo i e al., 2009]. Consis en wi h such a gumen s, bo h he excess ca bon expo (+Fe minus Fe ca bon expo ) and excess Fe inpu (+Fe minus Fe Fe inpu ) o IBIS we e abou hal ha o CROZEX (552 e sus 1041 mmol C m 2 and 0.031 e sus 0.062 mmol Fe m 2 , espec i ely). As ou lined abo e, ou es ima es o C e x a e, by defini ion, based on new Fe inpu s o he uppe ocean. Howe e , a la ge ac ion o he o al Fe used by phy oplank on is likely de i ed om egene a ion, o example h ough g azing and i al lysis [Boyd e al., 2012, 2005; S zepek e al., 2005]. Ou simple explana ion o he appa en ela ionship be ween C e (x) and bloom du a ion (Figu e 3), as dic a ed by he pe iod o e which Fe is e ained and/o esupplied o he sys em will, in eali y, likely eflec a ange o complex mic obial ecycling p ocesses in he mixed laye [Boyd and Ellwood, 2010; S zepek e al., 2005]. Subsequen consequences o C e x ,asdefined, will depend on he ex en o which Fe o C/mac onu ien s a e p e e en ial emine alized wi hin a sys em [F ew e al., 2006; Twining e al., 2014]. Fo example, i C is e ec i ely emine alized a shallowe dep h han Fe, po en ially due o ( e-)sca enging o he la e [F ew e al., 2006; Twining e al., 2014], C e x migh be u he dep essed as he a io be ween new and ecycled Fe supply ( e med he “ e- a io”[Boyd e al., 2005]) dec eases, as migh be expec ed in ch onically low Fe sys ems o he la e s ages o he seasonal cycle [Tagliabue e al., 2014]. O e all, on he basis o a ailable da a (Figu e 3), i hus appea s ha C e exhibi s sys ema ic a iabili y which is no a simple unc ion o he magni ude o new Fe inpu s (Figu e 3a) bu is ins ead ela ed o he mode o new Fe supply and he du a ion o he bloom/expo phase (Figu es 3b, 3c, and 3d), po en ially mode a ed u he by he in ensi y o subsequen ecycling. Consequen ly, di e ences in he na u al modes o new Fe supply likely esul in spa io- empo al g adien s in C e x wi hin bo h he mode n and paleo-oceans. Fo example, ela i ely low bu con inuous ben hic Fe inpu s o coas al upwelling egions would be expec ed o esul in high C e x ; howe e , his e ficiency would be expec ed o dec ease as inpu s inc ease, pa icula ly i i exceeds biological demand o he complexing capaci y o na u al o ganic Fe binding ligands [El od e al., 2004]. In con as , ae osols inpu s o Fe-limi ed egions, ep esen ing localized and seasonally a iable sou ces o Fe deposi ed o e imescales o weeks o mon hs [Jickells e al., 2005; Mahowald e al., 2009], migh be expec ed o be associa ed wi h mid ange alues o C e x , which could a y as a unc ion o deposi ion in ensi y, alongside o he ac o s which may al e o e all biological demand, such as he a ailabili y o mac onu ien s. A mo e comple e mechanis ic unde s anding o linkages be ween a mosphe ic CO 2 and pas [Ma inez-Ga cia e al., 2014; Ridgwell and Wa son, 2002] o u u e [Jickells e al., 2005] na u al Fe e iliza ion om dus deposi ion may hus need o conside such p ocesses. Mo e b oadly, any a emp s o quan i a i ely link he oceanic Fe and C cycles will need o conside he po en ial o deli e y mode o con ol he amoun o ca bon seques e ed pe uni Fe added. Re e ences Ach e be g, E. P., e al. (2013), Na u al i on e ilisa ion by he Eya jallojokull e up ion, Geophys. Res. Le .,40, 921–926, doi:10.1002/g l.50221. Bha ia, M. P., E. B. Kujawinski, S. B. Das, C. F. B eie , P. B. Hende son, and M. A. Cha e e (2013), G eenland mel wa e as a significan and po en ially bioa ailable sou ce o i on o he ocean, Na . Geosci.,6, 274–278. Blain, S., e al. (2007), E ec o na u al i on e iliza ion on ca bon seques a ion in he Sou he n Ocean, Na u e,446(7139), 1070–1074. Blain, S., G. Sa hou, and P. Laan (2008), Dis ibu ion o dissol ed i on du ing he na u al i on- e iliza ion expe imen KEOPS (Ke guelen Pla eau, Sou he n Ocean), Deep Sea Res., Pa II,55(5–7), 594–605. Bowie, A. R., M. T. Maldonado, R. D. F ew, P. L. C oo , E. P. Ach e be g, R. F. C. Man ou a, P. J. Wo s old, C. S. Law, and P. W. Boyd (2001), The a e o added i on du ing a mesoscale e ilisa ion expe imen in he Sou he n Ocean, Deep Sea Res., Pa II,48(11–12), 2703–2743. Boyd, P. W., and M. J. Ellwood (2010), The biogeochemical cycle o i on in he ocean, Na . Geosci.,3, 675–682. Boyd, P. W., e al. (2000), A mesoscale phy oplank on bloom in he pola Sou he n Ocean s imula ed by i on e iliza ion, Na u e,407(6805), 695–702. Boyd, P. W., e al. (2004), The decline and a e o an i on-induced suba c ic phy oplank on bloom, Na u e,428(6982), 549–553. Boyd, P. W., e al. (2005), FeCycle: A emp ing an i on biogeochemical budge om a mesoscale SF(6) ace expe imen in unpe u bed low i on wa e s, Global Biogeochem. Cycles,19, GB4S20, doi:10.1029/2005GB002494. Boyd, P. W., e al. (2007), Mesoscale i on en ichmen expe imen s 1993–2005: Syn hesis and u u e di ec ions, Science,315(5812), 612–617. Acknowledgmen s We hank all he scien is s and c ew membe s on IBIS c uises D350/354 (R.R.S. Disco e y, 1962–2012, NERC). We acknowledge M. S inchcombe, B. Head and K. Pabo sa a, A. Mil on, and C. Ma say o echnical suppo . This wo k was unded by NERC (NE/E006833/1, E.P.A. & C.M.M.) and he EU (CalMa O, 215157, R.J.S., F.A.C.L.M). Da a a e held a he B i ish Oceanog aphic Da a Cen e, h p://www.bodc.ac.uk/. We would like o hank wo anonymous e iewe s and he edi o o p o iding cons uc i e commen s. The Edi o hanks Robe S zepek and an anonymous e iewe o hei assis- ance in e alua ing his pape . Geophysical Resea ch Le e s 10.1002/2014GL060308 LE MOIGNE ET AL. ©2014. The Au ho s. 4625 19448007, 2014, 13, Downloaded om h ps://agupubs.onlinelib a y.wiley.com/doi/10.1002/2014GL060308 by Spanish Coch ane Na ional P o ision (Minis e io de Sanidad), Wiley Online Lib a y on [28/11/2024]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License Boyd, P. W., e al. (2012), Mic obial con ol o dia om bloom dynamics in he open ocean, Geophys. Res. Le .,39, L18601, doi:10.1029/ 2012GL053448. Buccia elli, E., S. Blain, and P. T egue (2001), I on and manganese in he wake o he Ke guelen Islands (Sou he n Ocean), Ma . Chem.,73(1), 21–36. Buessele , K. O., M. P. Bacon, J. K. Coch an, and H. D. Li ings on (1992), Ca bon and ni ogen expo du ing he JGOFS No h A lan ic Bloom Expe imen es ima ed om 234 Th: 238 U disequilib ia, Deep Sea Res., Pa I,39(7–8), 1115–1137. Buessele , K. O., J. E. And ews, S. M. Pike, and M. A. Cha e e (2004), The e ec s o i on e iliza ion on ca bon seques a ion in he Sou he n Ocean, Science,304(5669), 414–417. Che e , F., G. Sa hou, E. Buccia elli, S. Blain, and A. R. Bowie (2010), An i on budge du ing he na u al i on e ilisa ion expe iemen KEOPS (Ke guelen Islands, Sou he n Ocean), Biogeosciences,7, 455–468. de Baa , H. J. W., e al. (2005), Syn hesis o i on e iliza ion expe imen s: F om he i on age in he age o enligh enmen , J. Geophys. Res.,110, C09S16, doi:10.1029/2004JC002601. El od, V. A., W. M. Be elson, K. H. Coale, and K. S. Johnson (2004), The flux o i on om con inen al shel sedimen s: A missing sou ce o global budge s, Geophys. Res. Le .,31, L12307, doi:10.1029/2004GL020216. F ew, R. D., D. A. Hu chins, S. Nodde , S. A. Sanudo-Wilhelmy, A. To a -Sanchez, K. Leblanc, C. E. Ha e, and P. W. Boyd (2006), Pa icula e i on dynamics du ing FeCycle in suban a c ic wa e s sou heas o New Zealand, Global Biogeochem. Cycles,20, GB1S93, doi:10.1029/ 2005GB002558. Gledhill, M., and K. N. Buck (2012), The o ganic complexa ion o i on in he ma ine en i onmen : A e iew, F on . Mic ob.,28,3–69. Hawkings, J. R., J. L. Wadham, M. T an e , R. Raiswell, L. G. Benning, P. J. S a ham, A. Teds one, P. Nienow, K. Lee, and J. Telling (2014), Ice shee s as a significan sou ce o highly eac i e nanopa icula e i on o he oceans, Na . Commun.,5, 3929, doi:10.1038/ncomms4929. Henson, S., S. C. Pain e , N. P. Holliday, M. S inchcombe, and S. L. C. Gie ing (2013), Unusual subpola No h A lan ic phy oplank on bloom in 2010: olcanic e ilisa ion o No h A lan ic Oscilla ion?, J. Geophys. Res. Oceans,118,1–10, doi:10.1002/jg c.20363. Jickells, T. D., e al. (2005), Global i on connec ions be ween dese dus , ocean biogeochemis y, and clima e, Science,308(67), 67–71. Le Moigne, F. A. C., R. J. Sande s, M. Villa-Al ageme, A. P. Ma in, K. Pabo sa a, H. Planque e, P. J. Mo is, and S. J. Thomalla (2012), On he p opo ion o ballas e sus non-ballas associa ed ca bon expo in he su ace ocean, Geophys. Res. Le .,39, L15610, doi:10.1029/ 2012GL052980. Le Moigne, F. A. C., S. A. Henson, R. J. Sande s, and E. Madsen (2013a), Global da abase o su ace ocean pa icula e o ganic ca bon expo fluxes om he 234 Th echnique, Ea h Sys . Sci. Da a Discuss.,6, 163–187, doi:10.5194/essdd-6-163-2013. Le Moigne, F. A. C., M. Villa-Al ageme, R. J. Sande s, C. M. Ma say, S. Henson, and R. Ga cia-Teno io (2013b), Expo o o ganic ca bon and biomine als de i ed om 234 Th and 210 Po a he Po cupine Abyssal Plain, Deep Sea Res., Pa I,72,88–101. Le i us, S. E. (1982), Clima ological a las o he wo ld ocean, NOAA P o essional Pape , US Go e nmen P in ing O fice, Washing on, D. C. Mahowald, N. M., e al. (2009), A mosphe ic I on Deposi ion: Global Dis ibu ion, Va iabili y, and Human Pe u ba ions, Annu. Re . Ma . Sci.,1, 245–278. Mai i, K., e al. (2012), In e calib a ion s udies o sho -li ed ho ium-234 in he wa e column and ma ine pa icles, Limnol. Oceanog . Me hods,10, 631–644. Ma ino , I., A. Gnanadesikan, J. L. Sa mien o, J. R. Toggweile , M. Follows, and B. K. Mignone (2008), Impac o oceanic ci cula ion on biological ca bon s o age in he ocean and a mosphe ic pCO 2 ,Global Biogeochem. Cycles,22, GB3007, doi:10.1029/2007GB002958. Ma in, J. H., R. M. Go don, and S. E. Fi zwa e (1990), I on in An a c ic wa e s, Na u e,345(6271), 156–158. Ma inez-Ga cia, A., D. M. Sigman, H. Ren, R. F. Ande son, M. S aub, D. A. Hodell, S. L. Jacca d, T. I. Eglin on, and G. H. Haug (2014), I on e iliza ion o he Suban a c ic Ocean du ing he las ice age, Science,343, 1347–1350. Mo is, P. J., and M. A. Cha e e (2013), A syn hesis o uppe ocean ca bon and dissol ed i on budge s o he Sou he n Ocean na u al i on e iliza ion, Deep Sea Res., Pa II,90, 147–157. Mosse i, J., B. Queguine , L. A mand, and V. Co ne -Ba haux (2008), Impac o i on on silicon u iliza ion by dia oms in he Sou he n Ocean: A case s udy o Si/N cycle decoupling in a na u ally i on-en iched a ea, Deep Sea Res., Pa II,55(5–7), 801–819. Nielsdo i , M., C. M. Moo e, E. P. Ach e be g, R. J. Sande s, and D. J. Hinz (2009), I on limi a ion o h epos bloom phy oplank on communi ies in he Iceland Basin, Global Biogeochem. Cycles,23, GB3001, doi:10.1029/2008GB0034. Pain e , S. C., S. A. Henson, A. Fo yan, S. S eigenbe ge , J. Kla , M. S inchcombe, N. Rogan, A. Bake , E. P. Ach e be g, and C. M. Moo e (2014), An assessmen o he e ical di usi e flux o i on and o he nu ien s o he su ace wa e s o he subpola No h A lan ic Ocean, Biogeosciences,11, 2113–2130. Pike, S. M., K. O. Buessele , J. And ews, and N. Sa oye (2005), Quan ifica ion o Th-234 eco e y in small olume sea wa e samples by induc i ely coupled plasma-mass spec ome y, J. Radioanal. Nucl. Chem.,263(2), 355–360. Planque e, H., e al. (2007), Dissol ed i on in he icini y o he C oze Islands, Sou he n Ocean, Deep Sea Res., Pa II,57, 1999–2019. Polla d, R. T., e al. (2009), Sou he n Ocean deep-wa e ca bon expo enhanced by na u al i on e iliza ion, Na u e,457(7229), 577–U581. Poul on, A., A. Cha alampopoulou, J. R. Young, G. A. Ta an, M. I. Lucas, and G. D. Qua ly (2010), Coccoli hopho e dynamics in non-bloom condi ions du ing la e summe in he cen al Iceland Basin (July-Augus 2007), Limnol. Oceanog .,4(55), 1601–1613. Ridgwell, A. J., and A. J. Wa son (2002), Feedback be ween aeolian dus , clima e and a mosphe ic CO2 in glacial ime, Paleoceanog aphy,17(4, 1059), doi:10.1029/2001PA000729. Rijkenbe g, M. J. A., S. S eigenbe ge , C. F. Powell, H. an Ha en, M. D. Pa ey, A. R. Bake , and E. P. Ach e be g (2012), Fluxes and dis ibu ion o dissol ed i on in he eas e n (sub-) opical No h A lan ic Ocean, Global Biogeochem. Cycles,26, GB3004, doi:10.1029/2011GB004264. Ryan-Keogh, T. J., A. I. Macey, M. C. Nielsdo´ i , M. I. Lucas, S. S. S eingenbe ge , M. C. S inchcombe, E. P. Ach e be g, T. S. Bibby, and C. M. Moo e (2013), Spa ial and empo al de elopmen o phy oplank on i on s ess in ela ion o bloom dynamics in he high la i ude No h A lan ic Ocean, Limnol. Oceanog .,58(2), 533–545. Sande s, R., and T. D. Jickells (2000), To al o ganic nu ien s in D ake passage, Deep Sea Res., Pa I,47, 997–1014. Sande s, R., L. B own, S. Henson, and M. Lucas (2005), New p oduc ion in he I minge Basin du ing 2002, J. Ma . Sys .,55(3–4), 291–310. Sande s, R., P. J. Mo is, M. S inchcombe, S. Seeya e, H. Venables, and M. Lucas (2007), New p oduc ion and he a io a ound he C oze Pla eau in aus al summe 2004–2005 diagnosed om seasonal changes in ino ganic nu ien le els, Deep Sea Res., Pa II,54(18–20), 2191–2207. Sa mien o, J. L., and J. C. O (1991), 3-dimensional simula ions o he impac o Sou he n Ocean nu ien deple ion on a mosphe ic CO 2 and ocean chemis y, Limnol. Oceanog .,36(8), 1928–1950. Sme acek, V., e al. (2012), Deep ca bon expo om he Sou he n Ocean i on- e ilized dia om bloom, Na u e,487, 313–319. S zepek, R. F., M. T. Maldonado, J. L. Higgin, J. Hall, K. Safi,S.W.Wilhelm,andP.W.Boyd(2005),Spinning he"Fe ousWheel": The impo ance o he mic obial communi y in an i on budge du ing he FeCycle expe imen , Global Biogeochem. Cycles,19, GB4S26, doi:10.1029/2005GB002490. Geophysical Resea ch Le e s 10.1002/2014GL060308 LE MOIGNE ET AL. ©2014. The Au ho s. 4626 19448007, 2014, 13, Downloaded om h ps://agupubs.onlinelib a y.wiley.com/doi/10.1002/2014GL060308 by Spanish Coch ane Na ional P o ision (Minis e io de Sanidad), Wiley Online Lib a y on [28/11/2024]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License S zepek, R. F., K. A. Hun e , R. D. F ew, P. J. Ha ison, and P. W. Boyd (2012), I on-ligh in e ac ions di e in Sou he n Ocean phy oplank on, Limnol. Oceanog .,57(4), 1182–1200, doi:10.1029/2005GB002490. Sunda, W. G., and S. A. Hun sman (1995), I on up ake and g ow h limi a ion in oceanic and coas al phy oplank on, Ma . Chem.,50, 189–206. Tagliabue, A., J. B. Sallee, A. R. Bowie, M. Le y, S. Swa , and P. D. Boyd (2014), Su ace-wa e i on supplies in he Sou he n Ocean sus ained by deep win e mixing, Na . Geosci.,7, 314–320. Twining, B. S., S. B. Baines, N. S. Fishe , and M. R. Land y (2004), Cellula i on con en s o plank on du ing he Sou he n Ocean I on Expe imen (SOFeX), Deep Sea Res., Pa I,51, 1827–1850. Twining, B. S., S. D. Nodde , A. L. King, D. A. Hu chins, G. R. LeClei , J. M. De B uyn, E. W. Maas, S. Vog , S. W. Wilhelm, and P. W. Boyd (2014), Di e en ial emine aliza ion o majo and ace elemen s in sinking dia oms, Limnol. Oceanog .,59, 689–704. Geophysical Resea ch Le e s 10.1002/2014GL060308 LE MOIGNE ET AL. ©2014. The Au ho s. 4627 19448007, 2014, 13, Downloaded om h ps://agupubs.onlinelib a y.wiley.com/doi/10.1002/2014GL060308 by Spanish Coch ane Na ional P o ision (Minis e io de Sanidad), Wiley Online Lib a y on [28/11/2024]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License