Pervasive Iron Limitation at Subsurface Chlorophyll Maxima of the California Current Shane L

Home , Picoeukaryote

Pervasive iron limitation at subsurface chlorophyll maxima of the California Current Shane L. Hoglea,b,1, Christopher L. Dupontc, Brian M. Hopkinsond, Andrew L. Kinge, Kristen N. Buckf, Kelly L. Roeg, Rhona K. Stuarth, Andrew E. Allena,c, Elizabeth L. Manni, Zackary I. Johnsonj, and Katherine A. Barbeaua,1

aScripps Institution of Oceanography, University of California, San Diego, La Jolla, CA 92093; bDepartment of Civil and Environmental Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139; cMicrobial and Environmental Genomics, J. Craig Venter Institute, La Jolla, CA 92037; dDepartment of Marine Sciences, University of Georgia, Athens, GA 30602; eSection for Marine Biogeochemistry and Oceanography, Norwegian Institute for Water Research, NO-5006 Bergen, Norway; fCollege of Marine Science, University of South Florida, Tampa, FL 33620; gDepartment of Chemistry and Biochemistry, The College at Brockport, State University of New York, Brockport, NY 14420; hBiosciences and Biotechnology Division, Lawrence Livermore National Laboratory, Livermore, CA 94550; iTrace Metal Biogeochemistry Laboratory, Bigelow Laboratory for Ocean Sciences, East Boothbay, ME 04544; and jMarine Laboratory and Biology Department, Duke University, Beaufort, NC, 28516

Edited by David M. Karl, University of Hawaii, Honolulu, HI, and approved November 7, 2018 (received for review August 1, 2018)

Subsurface chlorophyll maximum layers (SCMLs) are nearly ubiq- (11). Although Fe/light colimitation has been observed in some uitous in stratified water columns and exist at horizontal scales high-latitude SCMLs (12, 13), mid-/low-latitude SCMLs from ranging from the submesoscale to the extent of oligotrophic both coastal and pelagic zones remain understudied. Dissolved gyres. These layers of heightened chlorophyll and/or phytoplank- Fe minima at SCMLs from the subtropical North Pacific gyre ton concentrations are generally thought to be a consequence of a (14) and the Sargasso Sea (15) may be a consequence of intense balance between light energy from above and a limiting nutrient biological demand, even Fe limitation, during summer months. flux from below, typically nitrate (NO3). Here we present multiple One study documented phytoplankton Fe/light colimitation from lines of evidence demonstrating that iron (Fe) limits or with light mesotrophic and oligotrophic SCMLs in the California Bight colimits phytoplankton communities in SCMLs along a primary and the eastern tropical North Pacific (16), while another found productivity gradient from coastal to oligotrophic offshore waters SCMLs in the oligotrophic Western Pacific to be mostly light in the southern California Current ecosystem. SCML phytoplank- limited with some groups of microbial eukaryotes potentially ton responded markedly to added Fe or Fe/light in experimental incubations and transcripts of diatom and picoeukaryote Fe stress Significance genes were strikingly abundant in SCML metatranscriptomes. Using a biogeochemical proxy with data from a 40-y time series, The vertical distribution of phytoplankton cells and chloro- we find that diatoms growing in California Current SCMLs are phyll concentrations throughout the sunlit water column is persistently Fe deficient during the spring and summer growing rarely uniform. In many ocean regions, chlorophyll concen- season. We also find that the spatial extent of Fe deficiency within California Current SCMLs has significantly increased over the last trations peak in distinct and persistent layers deep below 25 y in line with a regional climate index. Finally, we show that the surface called subsurface chlorophyll maximum layers diatom Fe deficiency may be common in the subsurface of major (SCMLs). SCML formation is hypothesized to reflect the con- upwelling zones worldwide. Our results have important implica- sequences of phytoplankton light/macronutrient colimitation, tions for our understanding of the biogeochemical consequences behavior, and/or photoacclimation. We discovered unexpect- of marine SCML formation and maintenance. edly persistent and widespread phytoplankton iron limitation and iron/light colimitation in SCMLs of the California Cur- nutrient limitation | iron | light | California Current | rent and at the edge of the North Pacific Subtropical Gyre deep chlorophyll maximum using shipboard incubations, metatranscriptomics, and biogeochemical proxies. These results suggest that interactions and feedbacks between iron and light availability play an e and light are essential for phytoplankton photosynthe- Fsis, but both resources are scarce in much of the ocean. important and previously unrecognized role in controlling the Surface ocean primary productivity is limited by the availabil- productivity and biogeochemical dynamics of SCMLs. ity of Fe in some regions (1), and mesoscale Fe fertilization experiments now firmly demonstrate that Fe availability con- Author contributions: S.L.H., C.L.D., B.M.H., K.N.B., E.L.M., Z.I.J., and K.A.B. designed research; S.L.H., C.L.D., B.M.H., A.L.K., K.N.B., K.L.R., R.K.S., E.L.M., Z.I.J., and K.A.B. per- trols phytoplankton biomass and growth rates in the Southern, formed research; S.L.H. contributed new reagents/analytic tools; S.L.H., C.L.D., B.M.H., equatorial Pacific, and subarctic Pacific oceans (2). In addition, A.L.K., K.N.B., E.L.M., Z.I.J., and K.A.B. analyzed data; and S.L.H., C.L.D., B.M.H., A.L.K., phytoplankton Fe limitation has been observed in midlatitude K.N.B., K.L.R., R.K.S., A.E.A., E.L.M., Z.I.J., and K.A.B. wrote the paper.y coastal upwelling zones (3, 4), throughout mesoscale circula- The authors declare no conflict of interest.y tion features (5, 6), and at the edge of subtropical gyres (7). This article is a PNAS Direct Submission.y In the surface ocean light attenuates rapidly to less than 1% of This open access article is distributed under Creative Commons Attribution-NonCommercial- incident photosynthetically available radiation (z1%) at depths NoDerivatives License 4.0 (CC BY-NC-ND).y from 50 m to 200 m, depending on turbidity. However, many Data deposition: All data supporting the findings of this study are available at the follow- diverse phytoplankton groups have adapted to growth at depths ing websites: CalCOFI time series data have been deposited in the CalCOFI data archives, approaching z1% despite the challenging low-light conditions. new.data.calcofi.org/index.php/reporteddata. Metatranscriptome and metagenome bio- Prior studies noting the overlapping scarcity of Fe and light in logical sequence files have been deposited in iMicrobe, https://imicrobe.us (iMicrobe much of the ocean predicted that these two resources synergisti- project ID: CAM P 0001069). Biogeochemical data subsets, metatranscriptome annota- cally colimit phytoplankton growth (8), particularly in subsurface tions, and all computer code required to reproduce the results reported in this study have been deposited in Zenodo, https://doi.org/10.5281/zenodo.1495558 and https://doi. chlorophyll maximum layers (SCMLs) (9). Indeed, work with org/10.5281/zenodo.1495504. Unprocessed biogeochemical data have been deposited in cultured phytoplankton demonstrates that Fe/light colimitation the UCSD Datazoo Research Project, oceaninformatics.ucsd.edu/datazoo/catalogs/ccelter/ can arise when demand for Fe-rich photosynthetic redox proteins sources/1758.y increases under low-light conditions (9, 10). However, the poten- 1 To whom correspondence may be addressed. Email: [email protected] or kbarbeau@ucsd. tial for phytoplankton Fe or Fe/light (co)limitation in SCMLs edu.y has been explored only in a handful of field studies despite the This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. significant feedbacks linking Fe/light (co)limitation, dust deposi- 1073/pnas.1813192115/-/DCSupplemental.y tion, and oceanic CO2 uptake in global biogeochemical models Published online December 10, 2018.

13300–13305 | PNAS | December 26, 2018 | vol. 115 | no. 52 www.pnas.org/cgi/doi/10.1073/pnas.1813192115 Downloaded by guest on September 25, 2021 Downloaded by guest on September 25, 2021 ol tal. et Hogle ovdF de ocnrtosadtecnetain fstrong, (L of ligands concentrations Fe-binding the organic and dis- concentrations total (dFe) measured Fe we offshore solved stations and three these transition, At this inshore, respectively. zones, on 93.3/80, the (20). (93.3/40, representing stations stations CC 93.3/120) CalCOFI three and southern standard sampled intensively the all we sampling of transect, to zones addition offshore) and In km inshore offshore), the km (∼450 spans (∼150–450 offshore which transition ), S1 offshore), transect Fig. km a Appendix, (∼0–150 along (SI factors 93.3 (Cal- line (co)limiting Investigations COFI) as Fisheries Oceanic light Cooperative investigated and California we of Fe 2007 of July During role 19). other the (4, phytoplankton and CC is the Fe-limited concentrations in generate communities waters Fe systems can in macronutrient-rich factors Current variation biogeochemical of Boundary local upwelling However, Eastern (18). intense other by and supported Fe (CC) phytoplankton rent of worldwide. prevalence SCMLs within the colimitation Fe/light establish scales or limitation to varied at needed approaches are experimental complementing mul- using tiple studies However, (17). colimitation Fe/light exhibiting cie(1m eesrnl eope tsain9./0(i.1A (Fig. 93.3/40 station of at decoupled (mean strongly layer were m) mixed (31 Fe acline surface (25). the colimitation nmol· than in 0.10 of rather low indicative limitation were were Fe and concentrations single SCMLs significant or inshore at to serial limitation pointed Sector. Fe consistently CC diatom experiments Inshore incubation the and (Si of proxies SCMLs cal at Limitation Fe Discussion culture and Results in 32). stress (31, field Fe the algal in and of 30) biosignatures (ISIP3), (29, 3 diagnostic protein on stress-inducible are iron focus known of which and we formerly 28), Here (27, (pTF), and ISIP2A (26). phytotransferrin as waters publication of surface prior patterns and a expression in SCML introduced from were were sequenced transcriptomes and Community limitation 11). collected serial section Appendix, or (SI colimitation, (25) simultane- mea- independent as these colimitation, responses incubation Using ous classified microscopy. then we by surements composition phytoplank- community and incubations, ton the NO of (Fv/Fm), course the measured II over photosystem We consumption of water. efficiency SCML quantum to tochemical light + Fe, Fe chlorophyll of combined additions and resource light, factorial with experiments incubation phytoplankton aggregate the incu- proxy’s of SCML the status zone emphasizing community. Fe transition further the and S3 ), to inshore Fig. responsiveness to , Appendix light (SI + bations Fe (SI and diatoms Fe Si for S2). conditions Fig. Fe-limiting Appendix, with consistent NO trations, high and Si (ρ low samples all across lated Appendix, Si (SI H deficiency. (24) of Fe deficiency uptake Fe diatom of Si preferential result Negative a 13). as section diatoms of tion Si Si the 4). section sup- indirect Appendix, or (SI direct the NO NO to phytoplankton attributed residual of be pression threshold can in this column nutrient water above than the colimiting greater or or is At limiting ratio 19). proximate N:dFe a the be where prox- waters to NO biogeochemical likely the is other deficiency: Fe two Fe determined for also ies We L 22). produce (21, bacte- actively associated may and/or incubation ria phytoplankton in that growth suggesting phytoplankton studies, Fe-limited with correlated ihpiaypoutvt ntesuhr aionaCur- California southern the in productivity primary High tec fteitnieysmldsain econducted we stations sampled intensively the of each At ∗ rx 2)adtae hfsi h lmna composi- elemental the in shifts traces and (23) proxy L a −1 ocnrtos rmr rdcinrts h pho- the rates, production primary concentrations, n h etso h ercie(4m n nitr- and m) (74 ferricline the of depths the and ) ex ex n :F ai) omnt transcriptomes, community ratio), N:dFe and n h :F ai eengtvl corre- negatively were ratio N:dFe the and 3 ex aesgnrlyas a o F concen- dFe low had also generally waters nrae usatal pnadto of addition upon substantially increased ex = ausi h ae ounindicate column water the in values ex 1 1 −0.75, sa dpaint elimitation Fe to adaptation an as (µmol·L .L ). 3 3 tlzto ylwF availability Fe low by utilization oderto(:F)adSi and (N:dFe) ratio dFe to 4 1 SiO ocnrtosaepositively are concentrations P 4 −1 .e7,idctn that indicating 2.4e-7), = eaiet NO to relative samdfidfr of form modified a is ) 3 ∼8 cuuainin accumulation o/ml(4, µmol/nmol Biogeochemi- 3 u to due ex 3 . .F ih tmltdtegets NO (15-fold greatest production the primary stimulated control), light over + (fourfold Fe drawdown (SI 11). colimitation Fe/light section independent Appendix, with consistent generate most N were to infrequent new too Si of persistent be a sources SCML. may and the growth important but diatom wind-stress zone Fe-limited likely chronic and euphotic are 34) the (35) this to (6, In upwelling features ). 13C.2 been curl circulation section have Appendix , mesoscale not (SI sector may elsewhere signal waters from Fe-limited the potentially advected and deficiency that Fe diatom indicating local from 1C), tran- (Fig. Si the zone negative of zone, most offshore in the the In deficiency 1B). Fe (Fig. diatom zone sition potential of region 8 a than Si greater ratios negative N:dFe and depth. acquisition this Fe at also of molecules production We biological enhanced uptake. with Fe L consistent biological localized a localized a observed enhanced measured poten- We to SCML, zone 1A). due transition outer tially (Fig. and the at 93.3/80 93.3/120 depletion Fe station station localized at at decoupled offset ferricline strongly moderately the were of depths nitracline layer the and and mixed respectively) surface zones, nmol·L offshore the and 0.13 in of low (mean again 93.3/120 colimitation. were Fe/light and levels concentrations potential 93.3/80 of light dFe of point stations signatures low the multiple at where to displayed SCMLs off- demand zone colimitation. Fe the euphotic photosynthetic Fe/light in the increased and have of SCML may and base zone nitracline the the transition of deepening toward the a Zones. observed Offshore of we zone and shore edge Transition oligotrophic CC the the at Fe Colimitation Fe/Light managing for pathways cellular investing pat- deficiency. into were expression phytoplankton of resources These SCML SCML significant inshore stress. the that Fe at suggest to terns transcripts due diatom also of expressed some 93.3/40—perhaps highly were (33), most diatoms the iron-limited for energy strategy Fe-independent alternative, acquisition an as proposed have been which recently genes, Proteorhodopsin-like 3). (Fig. proteins related ahtxnmcgopicuigrbsmlasml proteins, assembly from ribosomal functions cellular including NO essential group for taxonomic genes each identifiable expressed highly all of of dinoflagellates, and diatoms, transcripts. 1% by pelagophytes, community pTF/ISIP3 top of aggregated expression taxonomically the the Indeed, all belong- in of pTFs 1% were and top groups S5), Fig. these Phaeocystis Appendix, to groups taxonomi- (SI taxonomic ing 93.3/40 abundant all most at of the detected of percentile two 3). were rank (Fig. pelagophytes taxonomic transcripts 95th four community aggregated from the cally transcripts exceeding the with from 93.3/40 groups transcriptomes station pro- effect. and community other at no ISIP3 in to SCML families relative had S4). abundant functional Fig. alone strikingly tein Appendix, were light + with transcripts (SI Fe while and pTF conditions control) Fe Fe, incubation added over in and light The fivefold dominated diatoms light light. to chain-forming + Large (2.5- + Fe most Fe indistin- added the statistically and increased were Fe over rates conditions and added Fv/Fm 2 light (Fig. to in guishable added increase response and twofold in control chloro- approximately control total an in the experi- increase and incubation sevenfold tion, SCML to The five- 1B). phyll a (Fig. (N:dFe displayed N:dFe elevated ments of strikingly that rored were ratios 120 N:dFe SCML, and h ue rniinzn nuainrsossa 93.3/80 at responses incubation zone transition outer The σ 3 o/ml ihnteSM,adteSi the and SCML, the within µmol/nmol) .L S1). Table Appendix, SI θ rnpres etsoklk rtis n photosynthesis- and proteins, heat-shock–like transporters, a 6kg·m 26 = (Chl n eaoht rncit/sfrsa ela the as well as transcripts/isoforms pelagophyte and ,an a), PNAS ex −3 aushglgtdtebs fteSM as SCML the of base the highlighted values spca odph 0 eo h euphotic the below m 100 depths to isopycnal Phaeocystis | .-odices nbl NO bulk in increase ∼2.5-fold .Piayproduction Primary S3). Fig. Appendix, SI eebr2,2018 26, December −1 1 nraea h fsoezn SCML, zone offshore the at increase n .1nmol·L 0.11 and 1 fe iae recee that exceeded or rivaled often ocnrtospae tthe at peaked concentrations | o.115 vol. −1 ntetransition the in Phaeocystis | ex o 52 no. rfiemir- profile 3 µmol/nmol ex consump- ex inlat signal tracked | max 13301 and At = 3

EARTH, ATMOSPHERIC, ENVIRONMENTAL AND PLANETARY SCIENCES SCIENCES z dFe [nmol L-1] L [nmol L-1] z z SCML AC1 NO3 Fe 93.3/120 93.3/80 93.3/40 z 1% 93.3/120 93.3/80 93.3/40

0 Chl 50 1.2

50 a 0.9 [μg L 150 0.6 100 -1 0.3 ] 150 Depth [m] 250 200

50 Si

250 ex

[μmol L 5 0.25 0.75 1.25 0.51.0 1.5 0.51.0 1.5 B 0 93.3/120 93.3/80 93.3/40 150

-1 −5 ] 0 Depth [m]

50 250 24.5 24 100 24 50 NO 40 24.5 3

150 [μmol L 30 Depth [m] 25 20 200 150 25.5 10 -1

26 ] 0 250 26.5 -0.50.0 0.5 -0.30.0 0.3 -0.5 0.00.5 1.0 250 -1 a -1 Siex/10 [μmol L ] Chl [μg L ] 600 400 200 0 z N:dFe/100 [μmol/nmol] 1% Distance from Shore [Km]

Fig. 1. (A) Profiles of dissolved Fe (orange) and L1 (purple) concentrations. The green box denotes depths where chlorophyll a (Chl a) is within 50% of the concentration at the SCML depth (z ), the dashed black line is nitracline depth (z ), and the dashed red line is ferricline depth (zFe) ( , Table SCML NO3 SI Appendix S1 and section 4). (B) Profiles of Chl a (green), Siex/10 (blue), and N:dFe/100 (red). The black line represents the depth of 1% of incident irradiance (z1%). Red −3 (N:dFe ≥ 8), blue (Siex < 0), and purple (both proxies) indicate depth ranges with potential diatom Fe deficiency. (C) Chl a, NO3, and Siex (26.5 kg·m source isopycnal) sections. Potential density anomaly (σθ ) contours are in black. Small circles show nutrient sampling depths and large circles show incubation and metatranscriptome depths. White and red dashed lines show z1% and zSCML, respectively.

over control), and total Chl a (5.5-fold over control) increases, SCML Fe Deficiency Estimated from a 40-y Time Series. We sought while responses to Fe or light alone were similar to each other to characterize the potential for diatom Fe deficiency in the and significantly smaller than for Fe + light. However, the single southern CC across broader spatial and temporal scales by lever- addition of iron or light significantly enhanced NO3 drawdown aging 40 y of monthly sampling data collected from 75 stations (Fe, 1.7-fold over control; light, 1.5-fold over control) and the distributed over a 190,000-km2 area as a part of the CalCOFI pro- a addition of iron significantly increased the Chl concentration gram (calcofi.org/). We used Siex as a biogeochemical proxy for (2.5-fold over control). Large chain-forming diatoms dominated diatom Fe deficiency because H4SiO4 and NO3 measurements incubation responses to added Fe + light (SI Appendix, Fig. S4), are readily available in the CalCOFI dataset and there is a strong while iron-stress transcripts were in the top 1% of all expressed correlation between negative Siex and experimentally determined Phaeocystis and pelagophyte genes in situ (Fig. 3). Incubation Fe limitation in our results (Figs. 1 and 2 and SI Appendix, Fig. S3) experiments from 93.3/120 displayed a roughly twofold increase and the results of others (5, 6, 24). The Siex tracer assumes mini- in diatom cell numbers, NO3 drawdown, Chl a concentrations, mal effects from horizontal mixing/advection, shifts in upwelling and primary production in Fe + light conditions over the control source depth, and other processes (nitrification, denitrification, (Fig. 2). Only the increase in Chl a was statistically significant, variability in Si:N remineralization ratios) that may integrate non- but it was challenging to detect significant differences in the specific biogeochemical signals. In most settings, such as in the means across incubation replicates because responses in the euphotic zone of the CC, these assumptions appear to be valid, offshore zone were substantially smaller and noisier than in but we note that in other biogeochemical regimes they may not other regions. Given this caveat, our results suggest a poten- be. In all test cases and datasets we examined Siex performed as tial for simultaneous Fe/light colimitation in oligotrophic SCMLs a robust tracer (SI Appendix, section 13A). in the offshore zone. Smaller non–chain-forming diatoms were During the spring and summer months over the last 40 y the dominant responders to the offshore zone Fe addition in- at least 30% of all SCML samples in the southern CC were cubations (SI Appendix, Fig. S4), and the in situ positive Siex Fe deficient to some degree (Siex < 0). Fe deficiency is dis- signal at the offshore SCML suggests that heavily silicified, proportionately concentrated in SCMLs from the inshore (43% Fe-limited forming diatoms were not abundant at the time negative) and transition zones (26% negative) compared with of sampling. Indeed, pelagophytes numerically dominated the the offshore zone (7% negative) (Fig. 4 and SI Appendix, metatranscriptomes at 93.3/120, and pelagophyte ptF/ISIP3 tran- Fig. S6). On average Siex from inshore and transition zone scripts were in the top 1% of all expressed transcripts in the SCMLs has steadily become more negative since 2000 in con- metatranscriptome and in the top 0.1% of pelagophyte tran- trast to the general increase in the 1990s and most of the 1980s scripts indicating significant cellular resource investment into Fe (SI Appendix, Fig. S7). We also find that the total spatial area acquisition by small, nonsilicifying phytoplankton. of Fe-deficient SCMLs has significantly increased for the inshore

13302 | www.pnas.org/cgi/doi/10.1073/pnas.1813192115 Hogle et al. Downloaded by guest on September 25, 2021 Downloaded by guest on September 25, 2021 2,3,3) atclryfrteofhr n transition and , Appendix offshore Si (SI the nitra- Negative 2000s. for CC early the particularly the southern during 36), SCML, the zones 35, in the (20, zone S8) of euphotic Fig. in shoaling the occurred and general have trends cline, a These 4E). with (Fig. y conjunction 40 last the over sector control. addition no respectively. the 93.3/120, Chl for and measurement the 93.3/80, time-point in 93.3/40, line for horizontal h 216 dashed and The h, 168 h, incubations 120 of were lengths The detected). (not for limitation test silicate to diatom treatment incubation potential light + silicate a included we 93.3/120 At 0.05). pairwise and (ANOVA different statistically means have indistinguishable conditions statistically Different-colored same. are the that colored are means experi- group incubation with Conditions experiment of circles). each (individual point in incubations replicate time of final means are the Bars ments. from Fv/Fm and concentration), 2. Fig. ol tal. et Hogle of indicating extent potentially 4E), spatial (Fig. SCML inshore the the at deficiency in Fe diatom increase significant corre- Si a also and (39) negative with 1993 4 sponds since extreme (Fig. amplitude NPGO most y the of the 40 strengthening last with the in coincide from conditions and events upwelling-favorable (39) in associated CC result are the that S10) shifts Fig. (39). wind Appendix, CC with (SI southern intervals the in NPGO of fluctuations Positive hydrography with and regional correlated biogeochemistry a significantly CC. the is is southern mode that (NPGO) the index Oscillation in climate Gyre variability Pacific climate North of The modes Variables. decadal track Biogeochemical and Si lation, Negative of Patterns tracers other for observed been 38). has (37, as diatom Si zone inshore negative situ euphotic a generated offshore of in signal the propagation chronic along-isopycnal of it reflect by parts below may deepest caused and/or extended the be at or may deficiency S9 ) signal Fe Fig. This Appendix, 1A). (SI (Fig. Negative zone euphotic production. the primary of contribution coastal major Si to a suggesting diatoms S6 ), Fe-deficient Fig. , Appendix Chl (SI observed zone highest the with 0.0 0.2 0.4 80 40

0 ex 0 2 4 0 1 2 Control ausi h fsoezn eeal rce h aeof base the tracked generally zone offshore the in values

h ,Piaypouto P) NO (PP), production Primary a, Chl Light

93.3/120

Si+Light

Fe Fe+Light 100 0.0 0.2 0.4 0.0 2.5 5.0 7.5 50 0 0 2 4 ex

CWns h ot aicGr Oscil- Gyre Paciﬁc North the Winds, CC , t Control etwt as icvr rate-corrected discovery false with test

93.3/80 Light

a a Fe n vF aesi h eninitial mean the is panels Fv/Fm and ocnrtosi h inshore the in concentrations

3 rwon(ecnaeo initial of (percentage drawdown Fe+Light .The S7 ). Fig. Appendix, SI CLF deﬁciencymay Fe SCML ex 100 0.0 0.2 0.4 15 10 15 10 50

ausotncoincided often values

0 0 0 5 5 Control

93.3/40 Light

Fe Fe+Light

Chl Chl Fv/Fm ] L [µg PP PP

] d m C [mg NO [% drwdn] [% a

-1 -1 ≤ -3 ex ex UIC,rbls-,-ipopaecarboxylase. ribulose-1,5-bisphosphate RUBISCO, (Chl center proteins complex reaction light-harvesting and a/bc photosystem -W) -S, -R, are rel., PR-like, -H–N, -F, 99th ranking (PsbE, I/II transporter; proteins percentile nitrate the PS of NRT1, proteorhodopsin-like; exceed 10% proteins; eukaryotic line lower heat-shock HSP, the dashed included. in the not those 95th above and library-specific and ranking those taxonomic percentile while the exceed ranking lines percentile tran- black solid unique the transcripts/isoforms 10 Individual above displayed. than are for transcripts/isoforms greater values expression individual groups otherwise with boxplots; by functional represented groups Indi- are into station. scripts/isoforms million Functional each binned per at axis). are groups transcript (horizontal taxonomic transcripts/isoforms of axis four assembled kilobase the vertical of vidual per The each reads processes. to reads cellular the mapped essential represents to scale) related (log10 genes and genes 3. Fig. (ρ the depths SCML with and correlated nitracline the moderately between were Si difference SCMLs forcing. CC atmospheric southern of at patterns nals decadal and SCML Si the regional necessarily between not relationship mechanistic does a and exclude or responses (40) (41) nonlinear integrative reflect cumulatively actually may correlation weak apparent < (ρ weak quantitatively are rates between correlations time-lagged Si the However, forcing. shared ecec eosre tSMsfo h otenC,we CC, southern the from SCMLs at observed we deficiency Si shoaling isopycnal general. in increased strength and upwelling deficiency or Fe diatom with tion stoichiomet- mismatched Si a between NO reflect top of the may supply below which ric form nitracline, that the SCMLs with of associated is deficiency Fe between correlation positive (ρ Si The SCML ). S11 Fig. the Appendix, at (SI density 1e-200) potential the and 1e-200) ex ex Taxonomically resolved RPKM ex 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 xetadwn-tescr rEmndie osa upwelling coastal Ekman-driven or curl wind-stress and extent rmaGoa Perspective. Global a from n h ircieSM et fstsget htdiatom that suggests offset depth nitracline/SCML the and rfcxnhnbnig;RB,alrbsmlasml rti subunits; protein assembly ribosomal all RIBO, binding); fucoxanthin or 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3

eaieaudneo rncit/sfrsfo nw iron-stress known from transcripts/isoforms of abundance Relative pTF/ISIP3 Flavodoxin 331093.3/80 93.3/120 ex PR-like n CLptnildniymyrﬂc nassocia- an reﬂect may density potential SCML and

PNAS NRT1

HSP 3

RIBO correlation negative The below. from Fe and | RUBISCO eebr2,2018 26, December PS I/II rel. pTF/ISIP3 Flavodoxin PR-like Fe pervasive the contextualize To

NRT1 0.2; HSP

RIBO This S11). Fig. Appendix, SI | RUBISCO o.115 vol. PS I/II rel. pTF/ISIP3 Flavodoxin 93.3/40 | PR-like = o 52 no. NRT1 −0.49, 0.53, = HSP ex RIBO at trends |

RUBISCO ex 13303

PS I/II rel. sig- Dinoflagellates Diatoms Pelagophytes Phaeocystis < <

EARTH, ATMOSPHERIC, ENVIRONMENTAL AND PLANETARY SCIENCES SCIENCES A Jun/Jul 2004 1.00 C R from σ = 26.5 kg m−3 Offshore 34°N Si:N θ R from σ = 25.8 kg m−3 0.50 Si:N θ A B 32°N 0.00 1.00

Si at SCML Transition ex D -4 -2 0 2 4 A 30°N 0.50 B B Jun/Jul 2007 34°N 0.00 1.00 A

E Inshore 32°N B 0.50

Siex at SCML -4 -2 0 2 4 30°N 0.00 124°W 122°W 120°W 118°W 1980 1990 2000 2010

−1 Fig. 4. Time series (1977–2017) of Siex (µmol·L ), at the SCML in the CalCOFI sampling area. (A and B) Interpolated values of Siex at the depth of the SCML for (A) June/July 2004 (season with most negative SCML Siex values) and (B) June/July 2007 (approximate time of cruise). Small circles depict CalCOFI sampling stations, while large circles display locations of incubations and metatranscriptomes. Dashed lines separate inshore–transition–offshore zones. (C–E) Average spring–summer (April–September) fraction of SCMLs with negative Siex for each zone. A and B time points in C–E correspond to the areal contour plots in −3 −3 A and B. Red and blue series show two different Siex estimates using a source upwelling isopycnal of 26.5 kg·m and 25.8 kg·m , respectively. (E) There is a signiﬁcant monotonic increasing trend (nonparametric Mann–Kendall test P < 0.002) in the extent of negative Siex for the inshore region since 1990 (dashed lines are linear regressions for each source isopycnal). Shaded regions show when the NPGO index is positive.

also determined Siex at a global scale at 50 m depth using data can be Fe deficient, but we do observe a significant effect of Fe + from the 2013 World Ocean Atlas (42) (SI Appendix, section 6). light on total Chl a concentrations and notably high expression Our results highlight the subsurface of coastal upwelling zones of phytoplankton Fe-stress genes in situ. Based on our results as potential hotspots of diatom Fe deficiency (Fig. 5). The CC we speculate that oligotrophic SCMLs may experience periods (Figs. 4 and 5A) and the Humbolt/Peru Current (Fig. 5B) sys- of Fe/light colimitation or Fe serial/single limitation, but fur- tem show the largest negative Siex signals at 50-m depth, while ther direct observations in the oligotrophic ocean are needed to the Benguela and Canary Currents (Fig. 5 C and D) display confirm this hypothesis. negative signals to a lesser extent. The Oman upwelling zone Persistent diatom Fe (co)limitation at SCMLs likely has and the North Arabian Sea also appear to have notably Fe- downstream consequences for the carbon cycle. For example, deficient (Fig. 5E) Siex values in the subsurface. The greatest increased diatom silicification may enhance particulate carbon spatial extent of subsurface diatom Fe deficiency appears to export efficiency by increasing sinking rates and shielding cells overlap with well-known high-nutrient low-chlorophyll regions, from grazing in productive upwelling zones (5, 6). Furthermore, such as the Subarctic North Pacific (43) and the Pacific equa- historical biogeochemical patterns of diatom iron deficiency at torial upwelling zone (44). It is worth noting that many regions the CC subsurface appear to track dominant modes of cli- with negative Siex values identified in our analysis also over- mate variability in the North Pacific, which may be due to lie or potentially overlap with oxygen minimum zones (OMZs) regional atmospheric patterns that decouple the nitracline, the and/or anoxic OMZs (45) where denitrification and anammox ferricline, and the depth of the SCML. Biogeochemical mod- can decouple the cycling of H4SiO4 from NO3 and may impart els predict increased upwelling and NO3 fluxes to the southern a nonspecific signal to Siex. However, the occurrence of denitrification/anammox above the upwelling source isopycnal would counter any signal of iron-limited diatom uptake and push Siex in a positive direction. Phytoplankton iron (co)limitation has been Si at ex experimentally identified in surface waters from many of the neg- 50 m ative Siex regions identified here (7, 43, 44, 46, 47), but few studies A have investigated the potential for iron limitation at the depth 10 ranges of SCMLs in these regions. We find that subsurface waters C of many major oceanic upwelling regions display a biogeochem- E 0 ical imprint of diatom iron deficiency (negative Siex), which is B D consistent with our experimental and time series results from -10 SCMLs in the CC.

Conclusions The diagnosis of nutrient limitation in situ is a critical step ABCDE toward better understanding ﬂuxes of energy and matter in marine ecosystems. Our results suggest a strong coastal to offshore gradient in the combined effects of iron and light on SCML phytoplankton of the southern CC, a highly productive eastern boundary upwelling regime, and potentially upwelling −1 zones worldwide (Fig. 5). The shallower inshore and inner tran- Fig. 5. Global distribution of Siex (µmol·L ) at 50-m depth relative to an −3 ◦ sition zone SCML communities, which represent maxima in both upwelling source isopycnal of σθ = 26.5 kg·m . Data are from the 1 - diatom biomass and productivity, appear particularly susceptible resolved annual (1955–2012) mean isosurface 2013 World Ocean Atlas. A–E to Fe limitation or Fe/light colimitation. It is less obvious whether highlight mean negative Siex upwelling regions. The solid black line in the −3 deeper SCML communities from the oligotrophic offshore zones Southern Ocean shows the outcropping of the σθ = 26.5 kg·m isopycnal.

13304 | www.pnas.org/cgi/doi/10.1073/pnas.1813192115 Hogle et al. Downloaded by guest on September 25, 2021 Downloaded by guest on September 25, 2021 1 ikle ,Ocle 21)Ehne estvt foencC2utk odust to uptake CO2 oceanic of sensitivity Enhanced (2015) A Oschlies interactions L, Iron-light (2012) Nickelsen PW Boyd 11. PJ, Harrison RD, Frew KA, Hunter RF, Strzepek 10. 4 ol A egus A asrR,Mhwl 20)Io,mnaee n lead and manganese, Iron, (2005) N Mahowald RA, Kayser BA, irradi- Bergquist and EA, supply Boyle iron 14. by growth phytoplankton of Control (2001) al. et PW, Boyd 13. phytoplank- of Co-limitation (1999) NM Price PJ, Harrison PW, Boyd MT, Maldonado 12. 3 amet L rbrN reisiM,DneJ 20)Hg-aiuecnrl of controls High-latitude (2004) Copiotrophic JP (2016) Dunne MA, KA Brzezinski N, Barbeau Gruber EE, JL, Sarmiento Allen 23. JM, Blanton RM, Bundy SL, Hogle chem- iron 22. ambient on complexation organic the of in role The zone (1997) euphotic KW the Bruland EL, of Rue shoaling 21. Multi-decadal coastal (2009) California MD in Ohman macronutrients DL, and current Aksnes Iron boundary (2001) 20. eastern GJ Smith four EL, in Rue regimes KW, Bruland Production 19. (2003) EJ Kearns phytoplankton ME, on light-limitation Carr and 18. iron- of effect The (2010) al. limita- et light ZI, and Johnson iron 17. of influences Interactive (2008) KA Barbeau BM, Hopkinson study 16. time-series Atlantic (Bermuda sea Sargasso the in Iron (2005) al. et PN, Sedwick 15. 4 igA,BreuK 21)Dsovdio n arntin itiuin nthe in distributions macronutrient and iron Dissolved (2011) KA Barbeau AL, King 24. .Piaypoutvt n photo- and productivity Primary ). S3 Table Appendix, at (SI incubator Fe Percival added a and in housed conducted bottles, were 16 Ecological polycarbonate incubations 4-L Long-Term operating acid-cleaned duplicate standard Ecosystem in or the Current Triplicate program. following California Research analyzed the sampled and from were sampler procedures Macronutrients rosette (26). mea- methods) publication a were relevant prior all ligands using a (including Fe-binding data 24. in and transcriptomic introduced and The 4 were 50. 4 ref. ref. refs. in in in as as sured described measured as incubation was Fe processed and Dissolved and concentrations, collected were ligand experiments iron-binding concentrations, iron ehddtisaeaalbein available are details Method Methods and Materials currents. of boundary eastern SCMLs other in potentially productivity and CC primary the and may cycling climate changing carbon the influence This which by a proportionally. iron, by increase demonstrates limitation mediated not connection, Fe linkage do toward atmospheric–biogeochemical may fluxes SCML Fe potential which associated the 49), the at (48, if communities change diatom climate drive anthropogenic under CC ol tal. et Hogle .SnaW,Hnsa A(97 nerltdiflec fio,lgtadcl ieon size cell and light iron, of influence Interrelated (1997) SA Huntsman WG, Sunda as growth 9. phototrophic of efficiencies use Fe and Mn of to gyre. Predictions oceanic due (1990) an JA of export Raven boundary the carbon 8. at co-limitation enhance Nutrient (2017) fronts al. et ocean TJ, Mesoscale Browning (2017) 7. al. et MR, car- sustains Stukel stress 6. iron from ballasting silica Enhanced (2015) al. et MA, southern Brzezinski the in limitation 5. iron phytoplankton for Evidence ratios (2007) K uptake Barbeau Si:N AL, and King growth 4. diatom Iron-limited (1998) KW Synthesis Bruland 1993-2005: DA, experiments Hutchins enrichment 3. iron limitation. Mesoscale (2007) nutrient al. oceanic et of PW, patterns Boyd and 2. Processes (2013) al. et CM, Moore 1. eoiinb rnlgtcolimitation. iron-light by deposition phytoplankton. ocean Southern in differ growth. phytoplankton marine pathway. assimilation C of and growth for 116:1–18. availability light of function a Nature subduction. and sinking gravitational current. California the within zone frontal 120:4654–4669. a in export bon system. current California regime. upwelling coastal a in directions. future and Geosci tHwi ca iesre tto LH:Tmoa aiblt n nintermediate an and plume. variability hydrothermal Temporal water ALOHA: station time-series ocean Hawaii at project. SAZ the from results Res Experimental Geophys ocean: J Southern subantarctic the in ance ocean. Pacific subarctic NE Oceanogr the Stud in Top winter 2 during Part Fe Res and light by growth ton hroln uret n o aiuebooia productivity. biological latitude low and nutrients thermocline during production ligand iron-binding strong with blooms. phytoplankton associated addition are iron bacteria mesoscale marine a of response the and ocean Pacific experiment. equatorial the in istry system. current California the of sector southern blooms. diatom for Implications regimes: upwelling systems. ocean. Pacific western the of maxima 68:283–308. chlorophyll deep of communities Pacific. north eastern the in Oceanogr maxima Limnol chlorophyll subsurface at phytoplankton on tion ecological and variability, spatiotemporal imprint, Eolian implications. summer: during region) otenClfri urn system. current California southern ◦ iha1:2lgtdr yl tahg-adlwlgttreatment low-light and high- a at cycle light:dark 12:12 a with C 6:701–710. 551:242–246. epSaRsPr o tdOceanogr Stud Top 2 Part Res Sea Deep inlOceanogr Limnol lbBoece Cycles Biogeochem Glob 106:31573–31583. 53:1303–1318. Science inlOeng Lett Oceanogr Limnol a clPo Ser Prog Ecol Mar eci omci Acta Cosmochim Geochim 315:612–617. 46:2475–2485. 42:901–910. Nature Nature epy Res Geophys J IAppendix SI epy e Lett Res Geophys rcNt cdSiUSA Sci Acad Natl Proc 19:GB4006. 393:561–564. 390:389–392. inlOceanogr Limnol 342:91–103. 1:36–43. inlOceanogr Limnol 50:3199–3221. ape o oa dissolved total for Samples . 116:C03018. inlOceanogr Limnol 69:933–952. 42:492–499. 57:1182–1200. epy e Oceans C Res Geophys J 114:1252–1257. Nature 54:1272–1281. 46:1661–1674. 427:56–60. e Phytol New a Res Mar J epSea Deep Nat 7 oft W ta.(05 igohmsr fio nteAainsea. Arabian the in iron of Biogeochemistry (2015) al. et and JW, current Moffett Humboldt the 47. in limitation iron Phytoplankton (2002) al. et DA, oceanog- Hutchins Microbial 46. (2012) FJ Stewart RM, equatorial Letelier the EF, DeLong of DE, ecosystems Canfield in O, hypothesis Ulloa iron the 45. Testing (1994) al. phytoplank- et subarctic iron-induced JH, an Martin of fate 44. and decline The (2004) al. et PW, Boyd 43. (2014) ocean al. explain et to HE, hypothesis Garcia double-integration California. A 42. (2013) southern MD in Ohman and blooms E, climate algal Lorenzo Di ocean coastal Predicting links 41. (2017) oscillation al. gyre et Pacific JA, North McGowan (2008) 40. al. eastern et in production E, biological Lorenzo of Di reduction 39. Eddy-induced (2011) al. et N, California Gruber the in waters 38. upwelled of Resilience subduction the (2016) On R (2005) AW Goericke Mantyla pelagic SJ, JA, Bograd the Koslow MD, on 37. Ohman winds Jr, ocean DM Checkley of M, Influence Lindegren (2008) 36. Jr DM Checkley the RR, in Rykaczewski front deep-water 35. a to responses community Pelagic (2012) al. pro- et diatom MR, Marine Landry (2015) N 34. Cassar K, Ellis BM, Hopkinson D, Catlett A, Marchetti changes to 33. responses physiological and transcriptional Diatom of (2017) populations al. wild et in NR, Cohen limitation iron 32. of indicators Genetic (2014) al. et PD, Chappell 31. 0 ukK,BuadK 20)Tepyiohmclseito fdsovdio nthe in iron dissolved of speciation physicochemical The (2007) KW Bruland KN, Buck 50. current California the upwelling to supply coastal nutrient on Enhanced (2010) change JP climate Dunne RR, of Rykaczewski effects Anticipated 49. (2015) al. et A, Bakun 48. iron assessing for index molecular-based a of Development diatom (2017) pennate al. phytoplankton, the et A, of marine Marchetti response in 30. Whole-cell ubiquitous (2008) al. protein, et novel AE, A Allen (2015) 29. al. et J, Morrissey high-affinity controls 28. phytotransferrin Carbonate-sensitive (2018) productivity al. and et light JB, across McQuaid expression gene 27. and Genomes (2015) al. et CL, con- Dupont gradient 26. and factorial Bridging (2015) A Wacker D, Raubenheimer E, Sperfeld 25. eto 13A section Appendix, (SI a efre ne h upcso h SDEb arneLivermore Lawrence by DOE Biological US of the DE-AC52-07NA27344. Contract of Office work under auspices this the Laboratory Science, National the of and of under portion funds performed A Office was internal (DE-FC02-02ER63453). (DOE), Research JCVI Energy Environmental through and of funded Department were US and and Sam- Ocean expedition OCE-1026607, Global the pling from NSF/OCE-0417616, derived acknowl- are (LTER Metatranscriptomes K.A.B. Long-Term OCE-1637632). Program Z.I.J.). Ecosystem (to Current Research OCE-0550302 California 05-507098 the Ecological Grant Grant from NSF NSF support anonymous and additional by edges two E.L.M.) and funded and and sequencing, was K.A.B. code with work (to computer assistance staff Field with for the reviewers. assistance (JCVI) counts, manuscript for Data Institute cell Hackl Venter Oceanographic Thomas phytoplankton program Craig National Dr. for J. CalCOFI the Carter the the at Melissa at of Laboratory thank members We Climate to Center. Ocean gratitude the our and express We Horizon. ACKNOWLEDGMENTS. and CalCOFI 26. and 17 refs. from Atlas in described Ocean as World measured were physiology ai fH of ratio H 4 SiO euupwelling. Peru zones. minimum oxygen 16003. anoxic of raphy ocean. Pacific bloom. ton A Mishonov S, Levitus p. eds 25 76, MD), NESDIS Spring, Atlas Silver (NOAA, NOAA silicate). nitrate, (phosphate, ents forcing. climate to response ecosystem Ecology change. ecosystem systems. upwelling boundary current. ecosystem. marine pelagic a of stability and regions. upwelling in ecosystem study. A-front the 748. of Overview ecosystem: current availability. California iron low with coping in role 9:2745–2748. potential their and teorhodopsins provinces. ocean across bioavailability iron in oceanica Thalassiosira eigsa Alaska. sea, Bering model. system earth Lett an Res in stratification Geophys increased and warming global with ecosystem ecosystems. 60:1671–1688. diatoms. pennate bloom-forming in status tricornutum uptake. facilitates and surface cell the at iron concentrates diatoms. in uptake iron communities. microbial Pacific subtropical eastern in gradients consumers. to applied framework general Lett a Ecol Towards co-limitation: resource of cepts new.data.calcofi.org 4 ·L −1 4 98:1419–1433. a Res Mar J SiO 19:201–215. ] − Nature urCi hneRep Change Clim Curr oio starvation. iron to 4 o NO ([µmol Nature oNO to inlOceanogr Limnol PNAS 37:L21606. 63:863–885. 428:549–553. Lett Res Geophys inlOceanogr Limnol 371:123–129. yrgahcadntin aawr downloaded were data nutrient and hydrographic 3 rmtenrhatPcfi ocean. Pacific northeast the from at Nature etaktecpanadce fthe of crew and captain the thank We ol ca ta 03 o :Dsovdiogncnutri- inorganic Dissolved 4: Vol 2013, Atlas Ocean World | 3 σ tSM etswscluae sSi as calculated was depths SCML at ) ·L eebr2,2018 26, December and θ a Geosci Nat −1 58kg·m 25.8 = rcNt cdSiUSA Sci Acad Natl Proc 555:534–537. rcNt cdSiUSA Sci Acad Natl Proc ] h Si The https://www.nodc.noaa.gov. 47:997–1011. × 1:85–93. 35:L08607. 52:1800–1808. R Si:NO3 rcNt cdSiUSA Sci Acad Natl Proc 4:787–792. Phycol J rcRScBBo Sci Biol B Soc R Proc −3 ,weeR where ), rn a Sci Mar Front rcNt cdSiUSA Sci Acad Natl Proc r2. kg·m 26.5 or | 53:820–832. o.115 vol. 105:10438–10443. 105:1965–1970. SEJ ISME Si:NO3 urBiol Curr 4:360. lntnRes Plankton J −3 110:2496–2499. SEJ ISME 9:592–602. | stemicromolar the is 283:20151931. . o 52 no. inlOceanogr Limnol 25:364–371. Phaeodactylum 9:1076–1092. ex 109:15996– VNew RV ex [µmol = | 34:739– SEJ ISME proxy 13305

EARTH, ATMOSPHERIC, ENVIRONMENTAL AND PLANETARY SCIENCES SCIENCES