663

Journal of Oceanography, Vol. 59, pp. 663 to 670, 2003

Keywords:⋅ ENSO cycle (ElNiño/SouthernOscillation),

⋅ settling particles,⋅ organic carbon,⋅ carbonate,⋅ biogenic opal,⋅ amino acid,⋅ sediment trap,⋅ WPWP,⋅ equatorial Pacific.

* Corresponding author. E-mail: h.kawahata@aist.go.jp

Copyright © The Oceanographic Society of Japan.

El Niño/Southern Oscillation (ENSO) Related Variationsin Particulate Export Fluxes in the Western and CentralEquatorial Pacific

HODAKA KAWAHATA1,2* and LALLAN PRASAD GUPTA1

1Institute for Marine Resources and Environment, National Institute of Advanced Industrial Science and Technology (AIST), 1-1-1 Higashi, Tsukuba 305-8567, Japan2Graduate School of Science, Tohoku University, Aoba, Sendai 980-8578, Japan

(Received 29 June 2002; in revised form 12 September 2002; accepted 19 September 2002)

Time-series data from sediment trap moorings intermittently deployed during 1991–1999 show that the fluxes of biogenic material (carbonate, opal and organic matter,including amino acids) and other related parameters are temporally and spatiallydistinct across the Western Pacific Warm Pool (WPWP). These variations resultedfrom the El Niño and La Niña conditions, which alternately prevailed over the equa-torial Pacific Ocean during the mooring deployments. The westernmost WPWP (ahemipelagic region) recorded relatively high average total mass and amino acid fluxesduring the El Niño event. This was in sharp contrast to the eastern part of the WPWP(oligotrophic and weak upwelling regions) which recorded higher flux values duringthe La Niña event. Settling particulate organic matter was rich in labile components(amino acids) during La Niña throughout the study area. Relative molar ratios ofaspartic acid to β-alanine together with relative molar content of non-protein aminoacids (β-alanine and γ-aminobutyric acid) suggested that organic matter degradationwas more intense during La Niña relative to that during El Niño in the WPWP. Thisstudy clearly shows that during an El Niño event the well documented decrease inexport flux in the easternmost equatorial Pacific is accompanied by a significant in-crease in export flux in the westernmost equatorial Pacific Ocean.

The equatorial western Pacific, the Western PacificWarm Pool (WPWP), is characterized by the highest open-ocean water temperatures in the world (>28°C, Yan et al.,1992). The temperature and geographic extension of theWPWP fluctuate under the influence of ENSO (El Niño/Southern Oscillation) events. During El Niño yearsanomalously cool temperatures and high pressure developover the western equatorial Pacific, with low pressure andanomalously warm temperatures in the east. The tradewinds relax in the central and western Pacific, causingthe warm water in the western Pacific to spread eastwards,leading to depression of the thermocline in the easternPacific and preventing the upwelling of cold water offPeru. In contrast, totally opposite oceanographic condi-tions occur during La Niña events (Neelin et al., 1998).

In this paper we discuss the variations in biogeniccomponents of settling particles at a variety of WPWPstations. Amino acid (AA) based parameters, such asmolar ratios of aspartic acid and beta-alanine (Asp/β-Ala)and glutamic acid and gama-aminobutyric acid (Glu/γ-Aba), indicate the relative freshness or lability of settling

1. IntroductionAn important scientific consideration in relation to

the global carbon cycle is that atmospheric CO2 concen-tration is rapidly increasing due to human activities andit threatens to cause major changes in the global climate.Marine biogeochemical processes may play an importantrole in regulating the atmospheric concentration of CO2,through their mediation in air-sea exchange of CO2. Afraction of the biogenic material which is produced byphotosynthesis sinks or is actively exported out of thesurface waters into deeper water masses (Honjo, 1996).This process is referred to as the “biological pump”. Sincethe amount of carbon fixed by photosynthesis is 4–6 timesthat released by human activities, its fluctuations in re-sponse to climatic change are also a matter of great con-cern.

664 H. Kawahata and L. P. Gupta

particulate organic matter (OM); the higher the ratios thefresher the OM. Similar information is provided by thecontent of AA carbon relative to total organic carbon (AA-C%) and AA nitrogen relative to total nitrogen (AA-N%).Parameters like molar ratios of AA and hexosamine (AA/HA) suggest the relative contribution of zooplanktonicdebris to phytoplanktonic debris, because HA (mostlyglucosamine—Glc-NH2) is present in the exoskeleton ofzooplankton in the form of chitin. The source of HA canbe resolved based on the molar ratio of Glc-NH2/Gal-NH2(galactosamine), because this ratio is usually below 4 inthe bacterial biomass, but very high in the case ofzooplanktonic biomass. These biogeochemical parametershave been used in a number of studies dealing withparticulate OM in aquatic systems around the world (e.g.,Haake et al., 1992; Gupta et al., 1997; Jennerjahn et al.,

1999; Petersson and Floderus, 2001; Gupta and Kawahata,2003). The sampling sites were located in the hemipelagicregion off New Guinea of the western WPWP (Sites N1,N2 and M1), in the oligotrophic region of the central andeastern WPWP (Sites N3, N10 and M3) and in the equa-torial upwelling region of the eastern WPWP (Site M5).The data set is important with respect to the impact of ElNiño and La Niña events as these events might have in-fluenced settling particle flux and its biogeochemical na-ture.

2. Areas of Investigation and Oceanographic SettingThe study area was spread along the equator in the

WPWP up to its eastern margin (Fig. 1). Three major sur-face ocean currents—the north equatorial current (NEC),equatorial counter current (ECC) and south equatorial

Fig. 1. Seasonal variations in total mass fluxes (mg m–2day–1) at Sites N1, N2, N3, N10, M1, M3, and M5 in the Western PacificWarm Pool. Data for Sites N1, N2, N3 and N10 were obtained during El Niño events while those for Sites M1, M3, and M5were during La Niña event. Data for Sites N1, N2, N3 and N10 have already been published in Kawahata et al. (2000). Alsodata at Site N3 during 1990–91 (thin line) and 1992–93 (broken line) are cited from Harada et al. (1994).

ENSO Related Particle Flux Variations in the WPWP 665

current (SEC)—influence primary production and parti-cle flux around the sampling sites. West bound equato-rial currents (NEC and SEC) pile up warm water in theWPWP (Gordon and Fine, 1996). Besides these currents,the equatorial undercurrent (EUC), an eastbound flow,also has the potential to affect settling particle flux alongthe equator (Hansell et al., 1997). Based on modelingstudies of the equatorial Pacific, the oligotrophic upperlayer of the WPWP can be clearly distinguished, show-ing nitrate depletion and low new production, fromupwelled, saltier, cooler, nitrate-rich waters supportinghigh primary production in the eastern equatorial Pacific(Stoens et al., 1999).

In comparison with the eastern equatorial Pacific,the upwelling intensity is generally weaker in the west-ern equatorial Pacific. There is, however, a strongupwelling system called the Mindanao Dome in the west-ern equatorial Pacific (7°N, 130°E) (Lukas et al., 1991;Masumoto and Yamagata, 1991; Godfrey et al., 1993),where thermocline depth is much shallower than averagein the western equatorial Pacific (Peña et al., 1994). As aresult, the western sector of the WPWP shows primaryproduction as high as 600 mg C m–2day–1, while most ofthe central and eastern sectors of the WPWP show me-dium to low primary production of 200 to 400mg C m–2day–1 (Plates 1a–d in Antoine et al., 1996).

3. MethodologyMoorings consisting of three types of time-series

sediment traps (SMD21-6000, SMD26S-6000 andMark7G-21) were deployed for a period of about one yearin the WPWP. At Sites N1 and N2, moorings were de-ployed during 1991–92, at Site N3 during 1992–93, atSite N10 during 1994–95, and at Sites M1, M3 and M5during 1999. Prior to deployment, sampling bottles in eachtrap were filled with filtered seawater mixed with ana-lytical grade formalin to make a 3% solution (bufferedwith sodium borate) to retard microbial activity in thetrapped sample. Samples were collected at 2 week inter-vals. After recovery, the samples were stored at 2–4°C.In the laboratory they were passed through a 1 mm sieveto remove zooplanktonic swimmers. The

666 H. Kawahata and L. P. Gupta

Tab

le 1

. L

ocat

ions

of

the

trap

s, w

ater

dep

ths,

mea

n fl

uxes

, av

erag

e co

mpo

siti

ons

and

vari

ous

rati

os o

f m

ater

ial

coll

ecte

d at

diff

eren

t sit

es. S

ettl

ing

part

icle

s at

N1,

N2,

N10

and

N3

wer

e ta

ken

duri

ng E

l Niñ

o ev

ents

whi

le th

ose

at M

1, M

3 an

d M

5 w

ere

obta

ined

dur

ing

La

Niñ

a ev

ents

. Maj

or c

ompo

nent

s of

the

data

dur

ing

El N

iño

even

ts c

ome

from

Kaw

ahat

a et

al.

(19

98a)

and

Kaw

ahat

a et

al.

(20

00)

and

amin

o ac

id a

nd h

exos

amin

e da

ta a

re c

ited

fro

m G

upta

and

Kaw

ahat

a (2

000)

and

Gup

ta a

ndK

awah

ata

(200

2).

ENSO Related Particle Flux Variations in the WPWP 667

were nearly 2 to 3 times higher during 1991–92 (El Niño)compared to that (58.6 ± 55.4 mg m–2day–1 at Site M1)during 1999 (La Niña) in the hemipelagic region (Fig.3a). In the oligotrophic warm pool and weak upwellingregions, the situation was quite opposite. Both traps M3and M5 recorded higher fluxes (37.2 ± 31.1 and 53.7 ±

24.2 mg m–2day–1) relative to those (25.7 ± 10.5 and40.1 ± 28.5 mg m–2day–1) deployed at Sites N10 and N3,respectively (Kawahata et al., 1998b). This difference inthe average flux is statistically significant for the respec-tive regions; Student’s t-test, p = 0.01 (for Sites M3 andN10), 0.16 (for Sites M5 and N3), 0.07 (for Sites M3 andM5) and 0.16 (for Sites N10 and N3). Moderate fluctua-tion was identified in both regions in contrast to largefluctuation in the hemipelagic region.

The characteristics of the major biogenic componentsgenerally follow the patterns evident in the total massfluxes. It is evident from Figs. 3b–d that siliceous organ-isms are more prominent in the hemipelagic region rela-tive to the other two regions. Fluxes of carbonate, organicmatter (OM), biogenic opal and lithogenics were enhancedin the hemipelagic region during El Niño events. It hasbeen suggested that the COrganic/CCarbonate ratio of settlingparticles is an indicator of surface ocean fertility, increas-ing ratios reflecting a more efficient organic carbon pump(Berger and Keir, 1984). A greater contribution of bio-genic opal was associated with higher COrganic/CCarbonateratio, which led to the rise in the efficiency of the organiccarbon pump in this region. In contrast, the oligotrophicwarm pool and weak upwelling regions showed lowerfluxes of major components with little compositionalchange. The organic carbon pump was thus less efficientin these regions. The M3 trap occasionally recorded nearlyzero flux during fall (September–October) in theoligotrophic warm pool region. With respect to AA fluxvariations, the hemipelagic trap recorded much lowerfluxes during La Niña (Site M1) than those during El Niño(Site N2). However, in the oligotrophic warm pool andweak equatorial upwelling regions, the fluxes wereslightly higher during the La Niña event than those dur-ing El Niño.

Normalization of amino acid concentrations to OC(AA/OC) facilitates comparison among traps by alleviat-ing biases due to the inorganic components of the set-tling particles (Cowie and Hedges, 1992). Throughout theWPWP, settling particulate OM had more AA during LaNiña than during El Niño. AA carbon and nitrogen con-tents relative to OC and TN (AA-C% and AA-N%, re-spectively) were higher during La Niña relative to thoserecorded during El Niño. In addition, the Glc-NH2/Gal-NH2 mole ratio showed much larger values during La Niñain the oligotrophic warm pool and weak upwelling re-gions than during El Niño, but in the hemipelagic regiondifference between both the periods was relatively small.In the case of AA/HA mole ratio, the M1, M3 and M5traps recorded slightly higher values than the other traps.These five parameters generally showed a vertical de-crease between the shallow and deep traps due to the de-composition of settling particulate OM (Gupta andKawahata, 2002).

Fig. 3. Average fluxes of (a) total mass and carbonate, (b) or-ganic matter (OM), biogenic opal and lithogenics with (c)annual mean relative contribution, (d) biogenic opal/car-bonate (�) and COrganic/CCarbonate ratios (�). (e) Total aminoacid (AA) fluxes at various sites in the WPWP. Averagecontents of (f) amino acid carbon (AA-C%) (�) relative toOC and amino acid nitrogen (AA-N%) (�) relative to totalnitrogen (TN) in the WPWP. Average values of diageneticindicator ratios (g) amino acids and hexosamines (AA/HA)(�) and glucosamine and galactosamine (Glc-NH2/Gal-NH2) (�), (h) aspartic acid and β-alanine (Asp/β-Ala) (�)and glutamic acid and γ-aminobutyric acid (Glu/γ-Aba) (�).

668 H. Kawahata and L. P. Gupta

The enzymatic degradation of AA such as asparticacid (Asp) and glutamic acid (Glu) results in increasedconcentrations of non-protein amino acids β-alanine (β-Ala) and γ-aminobutyric acid (γ-Aba), respectively (Leeand Cronin, 1982). Asp/β-Ala values (mole ratios) werelower during La Niña than during El Niño in bothhemipelagic and weak upwelling regions, but showed lit-tle change at Sites M3 and N10 in the oligotrophic warmpool region, where Site N3 showed relatively higher val-ues for this ratio. Glu/γ-Aba values (mole ratios) werealso low in the weak upwelling area during La Niña, butin other regions it was nearly equal during El Niño andLa Niña. On the other hand, β-Ala + γ-Aba relative molecontent showed higher values throughout the study areaduring La Niña. These results indicate rapid degradationof settling OM during La Niña relative to that during ElNiño.

5. Discussion: Settling Particles in Response to ENSOFactors such as surface currents, tropical instability

wave (TIW), Kelvin wave, El Niño and La Niña, deter-mine the vertical transport of nutrients and amount of lightavailable to primary producers for photosynthesis in theequatorial Pacific. Under normal (non-El Niño and non-La Niña) conditions, the eastern edge of the WPWP, whereSite M5 was located, is characterized by weak equatorialupwelling, resulting in low to moderate primary produc-tivity (Halpern and Feldman, 1994; Gupta and Kawahata,2000). In contrast, in the western sector of WPWP aroundthe Indonesian maritime continent (Sites M1, N1 and N2),supply of nutrients through the Mindanao dome upwellingand terrestrial input due to tropical rainfall on the conti-nent support high primary productivity (Kawahata, 1999).Lying between these two regions, the oligotrophic warmpool region (Sites M3 and N10) has the lowest primaryproductivity (Antoine et al., 1996).

During an El Niño event this scenario changes de-pending upon the intensity of the event. Owing to weakeasterlies, cold equatorial water recedes in the east andthe warm water and rainfall zone around the Indonesianmaritime continent shift eastward during El Niño. Thisresults in a decreased supply of nutrients through surfacerunoff in the hemipelagic region, however, shoaling ofthe nutricline by 20–40 m was observed to enhance pri-mary productivity in surface water during the 1987 ElNiño (Radenac and Rodier, 1996). Reduction in primaryproduction and particle flux to the deep ocean take placein the weak upwelling region around the InternationalDateline due to diminished upwelling (Gupta andKawahata, 2000). However, the oligotrophic warm poolregion does not evidence any noticeable change in pri-mary production and particle flux during an El Niño event.These changes across the equatorial Pacific are in greatcontrast to what exists there during a La Niña event. The

onset of La Niña coincides with strong easterlies whichengender intense upwelling along the Peruvian coast, andthe cold water tongue arising out of it extends well be-yond the International Dateline where primary produc-tivity increases to higher than usual levels (Glantz, 1996).In the WPWP, higher than usual rainfalls supply enhancedamount of lithogenic matter and nutrients to thehemipelagic ocean. However, depression of nutricline alsooccurs at the same time due to accumulation of warmwater. This results in lower primary productivity and de-creases the particulate flux of biogenic components to thedeep ocean. The decrease in lithogenic flux is caused bya reduced biological pump.

In the hemipelagic region, fluxes of biogenic com-ponents, AA and lithogenics were reduced during La Niña,compared to the higher fluxes recorded (Sites N1 and N2)during El Niño. Shoaling of the nutricline is responsiblefor the higher fluxes during El Niño, when lithogenic fluxwas also enhanced. High AA-C%, AA-N%, and AA/HAratio indicated that the settling OM was more enriched inlabile components during the La Niña than during the ElNiño (Figs. 3e–g). In addition, a lower OM flux was as-sociated with less relative abundance of biogenic opalcontent during La Niña, which might be a result of a lessefficient organic carbon pump.

The oligotrophic warm pool region provides readilydistinguishable influence of the El Niño and La Niña con-ditions, especially when traps at Sites N3 and M5 arecompared. During an El Niño event, the Site N10 traprecorded lower fluxes of total mass, major componentsand AA (Figs. 3a, 3b and 3e). However, the Site N3 traprecorded higher fluxes. As evidenced by low AA-C%, AA-N%, and AA/HA ratio (Figs. 3f and 3g), the OM at SitesN3 and N10 was less labile compared to that at Site M3during La Niña. Although located in the same oligotrophicregion, the considerable distance (ca. 1600 km) betweenthe trap sites cannot be ignored as having some influenceon the parameters studied during the El Niño and La Niñachanges. Moreover, the trap at Site N10 was deployedduring a weaker El Niño event, while the higher flux re-corded at Site M3 may be due to its closer proximity tothe Mindanao Dome and may not be a reflection of ElNiño influence.

The weak equatorial upwelling region provides aclear picture of variations in fluxes and the labile natureof settling OM during a La Niña event. The trap mooringM5 was deployed at the same location as mooring N3was deployed during an El Niño event. Comparison ofdata from these two moorings showed that fluxes of totalmass, biogenic components and AA were higher duringLa Niña than during El Niño (Fig. 3). Although the bio-genic opal contribution was also higher during La Niñadue to the westward shift of the equatorial upwelling sys-tem, it made little difference to the organic carbon pump.

ENSO Related Particle Flux Variations in the WPWP 669

In the eastern equatorial Pacific (Site C; 1°N, 139°W), 2to 3 times more biogenic particle flux was observed dur-ing the 1984 La Niña than that during the 1983 El Niño(Dymond and Collier, 1988). This change in particle fluxwas attributed to a change in primary production due tothe impact of ENSO.

At Site N3, total mass flux and composition (car-bonate, OM, biogenic opal and lithogenics) showedprominent peaks between August and December 1992, andsmaller peaks during January and February 1993 (Fig. 1)(Kawahata et al., 2000). The timing of these peaks coin-cided with the cessation of westerly wind anomalies whichpersisted for the preceding months to the east of the Date-line at the equator. During the months August and Octo-ber 1992 and January 1993, the zonal wind condition re-turned to normal, and the equatorial thermocline shoaledto near-normal depths (Kessler and McPhaden, 1995).During the same period, SST cooled rapidly everywhereeast of the dateline in association with the strong thermalinstability wave activity. The changes in ocean surfaceforcing induced non-El Niño conditions favorable to en-hanced plankton growth, and consequently enhanced par-ticle flux.

Primary productivity under the El Niño condition hasbeen recorded as being lower than that under non-El Niñoconditions in the central and eastern equatorial Pacific(Chavez and Barber, 1987). Estimates show that the ElNiño conditions could decrease the carbon transport tothe interior of the equatorial Pacific by about 50%(Dymond and Collier, 1988). Similar to the observationmade in another study (Lee and Cronin, 1984), we no-ticed that variations in total mass flux are closely paral-leled by those in OM and labile components (AA) in ourstudy area, too. The OM is more labile during La Niñaand undergoes enhanced microbial degradation relativeto that during El Niño conditions.

AcknowledgementsThis study was supported by the “GCMAPS pro-

gram” (Global carbon cycle and related mapping basedon satellite imagery) promoted by Ministry of Education,Culture, Sports, Science and Technology of Japan. Fig-ures are drawn by Ms. K. Nagayoshi.

ReferencesAntoine D., J.-M. Andre and A. Morel (1996): Oceanic primary

production. 2. Estimation at global scale from satellite(coastal zone color scanner) chlorophyll . GlobalBiogeochemical Cycles, 10(1), 57–69.

Berger, W. H. and R. S. Keir (1984): Glacial-Holocene changesin atmospheric CO2 and the deep-sea record. p. 337–351.In Climate Processes and Climate Sensitivity, ed. by J. E.Hansen and T. Takahashi, Geophys. Monogr. 29, MauriceEwing Series, Vol. 5, AGU, Washington, D.C.

Chavez, F. P. and R. T. Barber (1987): An estimate of new pro-

duction in the equatorial Pacific. Deep-Sea Res., 34, 1229–1243.

Cowie, G. L. and J. I. Hedges (1992): Sources and reactivitiesof amino acids in a coastal marine environment. Limnol.Oceanogr., 37(4), 703–724.

Dymond, J. and R. Collier (1988): Biogenic particle fluxes inthe equatorial Pacific: Evidence for both high and low pro-ductivity during the 1982–1983 El-Niño. GlobalBiogeochemical Cycles, 2, 129–137.

Garrett, D. (2000): http//:www.cpc.ncep.noaa.gov/products/analysis_monitoring/bulletin/figt2.gif

Glantz, M. H. (1996): Currents of Change: El Niño’s Impacton Climate and Society. University Press, Cambridge, 194pp.

Godfrey, J. S., A. C. Hirst and J. J. Wilkin (1993): Why doesthe Indonesian Throughflow appear to originate from theNorth Pacific? J. Phys. Oceanogr., 23, 1087–1098.

Gordon, A. L. and R. A. Fine (1996): Pathways of water be-tween the Pacific and Indian oceans in the Indonesian seas.Nature, 379, 146–149.

Gupta, L. P. and H. Kawahata (2000): Amino acid and hex-osamine compositions and flux of sinking particulate mat-ter into the equatorial Pacific at 175°E longitude. Deep-SeaRes., 47, 1938–1960.

Gupta, L. P. and H. Kawahata (2002): Impact of ENSO vari-ability on the flux and composition of sinking POM in thewestern equatorial Pacific Ocean: Amino acids andhexosamines. Deep-Sea Res. II, 49, 2769–2782.

Gupta, L. P. and H. Kawahata (2003): Vertical and latitudinalvariations in fluxes and amino acid and hexosamine com-position of settling particulates along 175°E in the NorthPacific Ocean. Tellus B (in press).

Gupta, L. P., V. Subramanian and V. Ittekkot (1997): Biogeo-chemistry of particulate matter transported by the GodavariRiver, India. Biogeochemistry, 38(2), 103–128.

Haake, B., V. Ittekkot, V. Ramaswamy, R. R. Nair and S. Honjo(1992): Fluxes of amino acids and hexosamines to the deepArabian Sea. Mar. Chem., 40, 291–314.

Halpern, D. and G. C. Feldman (1994): Annual and interannualvariations of phytoplankton pigment concentration andupwelling along the Pacific equator. J. Geophys. Res.,99(C4), 7347–7354.

Hansell, D. A., N. R. Bates and C. A. Carlson (1997): Predomi-nance of vertical loss of carbon from surface waters of theequatorial Pacific Ocean. Nature, 386, 59–61.

Harada, K., S. Noriki, A. Shimamoto and N. Handa (1994):Latitudinal and seasonal change in sinking particles along175°E in the North Pacific. Kaiyo-Monthly, 26, 357–361(in Japanese).

Honjo, S. (1996): Fluxes of particles to the interior of the openoceans. p. 91–154. In Particle Flux in the Ocean, ed. by V.Ittekkot, P. Schäfer, S. Honjo and P. J. Depetris, SCOPE57, John Wiley & Sons Ltd., Chichester.

Jennerjahn, T. C., V. Ittekkot, C. E. V. Carvalho, A. R. C. Ovalle,C. E. Rezende and H. Erlenkeuser (1999): Temporal vari-ability of amino acid, hexosamine, and carbohydrate fluxeson the eastern Brazilian continental margin related to dis-charge of the São Francisco River, Brazil. Geo-Mar. Lett.,19, 202–208.

670 H. Kawahata and L. P. Gupta

Kawahata, H. (1999): Fluctuation in the ocean environmentwithin the western Pacific warm pool during LatePleistocene. Paleoceanogr., 14, 639–652.

Kawahata, H., M. Yamamuro and H. Ohta (1998a): Seasonaland vertical variations of sinking particle fluxes in the WestCaroline Basin. Oceanologica Acta, 21, 521–532.

Kawahata, H., A. Suzuki and H. Ohta (1998b): Sinking parti-cles between the equatorial and subarctic regions (0°N–46°N) in the central Pacific. Geochem. J., 32, 125–133.

Kawahata, H., A. Suzuki and H. Ohta (2000): Export fluxes inthe Western Pacific Warm Pool. Deep-Sea Res., 47, 2061–2091.

Kessler, W. S. and M. J. McPhaden (1995): The 1991–1993 ElNiño in the central Pacific. Deep-Sea Res. II, 42(2–3), 295–333.

Le Borgne, R., R. T. Barber, T. Delcroix, H. Y. Inoue, D. J.Mackey and M. Rodier (2002): Pacific warm pool and di-vergence: Temporal and zonal variations on the equator andtheir effects on the biological pump. Deep-Sea Res. II, 49,2471–2512.

Lee, C. and C. Cronin (1982): The vertical flux of particulateorganic nitrogen in the sea: decomposition of amino acidsin the Peru upwelling area and the equatorial Atlantic. J.Mar. Res., 40, 227–251.

Lee, C. and C. Cronin (1984): Particulate amino acids in thesea: Effects of primary productivity and biological decom-position. J. Mar. Res., 42, 1075–1097.

Lukas, R., E. Firing, P. Hacker, P. L. Richardson, C. A. Collins,R. Fine and R. Gammon (1991): Observation of the

Mindanao Current during the Western Equatorial PacificOcean Circulation Study. J. Geophys. Res., 96, 7089–7104.

Masumoto, Y. and T. Yamagata (1991): Response of the west-ern tropical Pacific to the Asian winter monsoon—the gen-eration of the Mindanao Dome. J. Phys. Oceanogr., 21,1386–1398.

Neelin, J. D., D. S. Battisti, A. C. Hirst, F. F. Jin, Y. Wakata, T.Yamagata and S. E. Zebiak (1998): ENSO theory. J.Geophys. Res., 103(C7), 14261–14290.

Peña, M. A., M. R. Lewis and J. J. Cullen (1994): New produc-tion in the warm waters of the tropical Pacific Ocean. J.Geophys. Res., 99(C7), 14255–14268.

Petersson, M. and S. Floderus (2001): Use of amino acid com-position to investigate settling and resuspension of a springbloom in the southern Skagerrak. Limnol. Oceanogr., 46(5),1111–1120.

Radenac, M. H. and M. Rodier (1996): Nitrate and chlorophylldistributions in relation to thermohaline and current struc-tures in the western tropical Pacific during 1985–1989.Deep-Sea Res. II, 43(4–6), 725–752.

Stoens, A., C. Menkes, M.-H. Radenac, Y. Dandonneau, N.Grima, G. Eldin, L. Memery, C. Navarette, J.-M. Andre, T.Moutin and P. Raimbault (1999): The coupled physical-newproduction system in the equatorial Pacific during the 1992–1995 El Niño. J. Geophys. Res., 104(C2), 3323–3339.

Yan, X.-H., C.-R. Ho, Q. Zheng and V. Klemas (1992): Tem-perature and size variabilities of the Western Pacific WarmPool. Science, 258, 1643–1645.