Coccolithophore export production, species composition, and coccolith-CaCO3 fluxes in the NE Atlantic ( and )

*Corresponding author. Fax: 020-646-2457.E-mail address: [email protected] (A.T.C. Broerse).

Deep-Sea Research II 47 (2000) 1877}1905

Coccolithophore export production,species composition, and coccolith-CaCO

3#uxes

in the NE Atlantic (343N 213W and 483N 213W)

Alexandra T.C. Broerse!,*, Patrizia Ziveri!, Jan E. van Hinte!,Susumu Honjo"

!Faculty of Earth Sciences, Vrije Universiteit, De Boelelaan 1085, 1081 HV Amsterdam, Netherlands"Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USA

Received 20 October 1997; received in revised form 15 December 1998; accepted 15 December 1998

Abstract

Coccolithophores were examined in two sediment traps, deployed for one year (April 1989 toApril 1990) at approximately 1 km water depth, in the subtropical NE Atlantic at 343N 213W(station NABE-34) and in the temperate NE Atlantic 483N 213W (station NABE-48) as part ofthe JGOFS North Atlantic Bloom Experiment (NABE). At both stations, maximum coccolithand coccosphere #uxes were recorded during local spring bloom periods. The average coccolithand coccosphere #uxes at NABE-34 were 1.12]109 m~2d~1 and 4.90]106 m~2d~1, respec-tively. At this station coccoliths from 51 taxa were identi"ed. A marked seasonal change wasnoticed in relative abundances of Emiliania huxleyi and deep photic zone species: E. huxleyidominated during spring, while in late autumn the coccolith assemblage was dominated byFlorisphaera profunda and the coccosphere assemblage by G. yabellatus and Algirosphaera spp..Emiliania huxleyi was the most abundant species in the overall coccolith and coccosphereassemblages with 69% and 64%, respectively. At NABE-48, the average daily #uxes weresigni"cantly lower, with 3.88]108 coccoliths m~2d~1 and 8.45]105 coccospheres m~2d~1,and only 36 taxa were recorded in the trap samples. Emiliania huxleyi was the dominant speciesin the coccolith assemblage throughout the year with 72%, and showed maximum #uxes duringspring 1990. Coccolithus pelagicus exhibited a distinct seasonal pattern, with maximum #uxes inearly June 1989, probably related to the development of a mesoscale cyclonic eddy in thevicinity of the trap site. Holococcolithophore #uxes were highest during maximum sea surfacetemperatures in summer. The coccosphere assemblage was dominated by Gephyrocapsa spp.,

0967-0645/00/$ - see front matter ( 2000 Elsevier Science Ltd. All rights reserved.PII: S 0 9 6 7 - 0 6 4 5 ( 0 0 ) 0 0 0 1 0 - 2

Syracosphaera spp., and E. huxleyi. Estimated CaCO3#uxes of coccoliths, coccospheres, and

calcareous dinophytes were compared with the CaCO3

content in the "ne ((32 lm) fraction.The annual mass estimated CaCO

3#ux of coccoliths and coccospheres was 5.1 g m~2 yr~1

at NABE-34 and 2.7 g m~2 yr~1 at NABE-48, and that of calcareous dinophytes 0.9 and0.1 g m~2 yr~1, respectively. These summed mass estimated values contributed on averageonly 55% at NABE-34 and 60% at NABE-48 to the "ne fraction CaCO

3. ( 2000 Elsevier

Science Ltd. All rights reserved.

1. Introduction

Coccolithophores are one of the major phytoplankton groups in the oceans.They a!ect the global climate system through the inorganic carbon pump (formationof coccoliths) and the organic carbon pump (photosynthesis), and by the emissionof dimethyl-sulphide (DMS) (Westbroek et al., 1993). Their calcite platelets(coccoliths) are major contributors to the carbonate in deep-sea sediments (Berger,1976; Milliman, 1993). The Atlantic is regarded as having the richest coccolithophore#ora (Gaarder, 1971), and the diversity is highest in low productivity regionsand areas of restricted circulation (Winter et al., 1994). Satellite imagery (CZCSand AVHRR) has shown that the ubiquitous species Emiliania huxleyi canform extensive blooms in the cooler, nutrient-rich waters of the NE Atlantic oceanbetween 503N and the southern coast of Iceland at about 633N (Holligan et al.,1993a; Brown and Yoder, 1994). Many studies have investigated the ecological andhydrographical factors that in#uence the location and the scale of these blooms(Tyrrell and Taylor, 1995; see in van Bleijswijk, 1996). Modeling results suggested thatthe most likely factors causing an E. huxleyi bloom are high light (high surfaceirradiances, shallow strati"cation) and low inorganic phosphate concentrations(Tyrrell and Taylor, 1995). Blooms of other coccolithophore species also havebeen reported in the temperate North Atlantic, such as Syracosphaera pulchra(Knappertsbusch and Brummer, 1995).

We selected two JGOFS sediment trap stations, in the subtropical (343N 213W)and the temperate (483N 213W) North Atlantic (Fig. 1) to conduct the "rstyear-round study on coccolithophore export production in sediment traps fromthe NE Atlantic at approximately 1000 m depth. We compared the coccolithophoresinking assemblages at the two distinct settings, the northern station close tothe potential bloom area of E. huxleyi, and the southern one in a moreoligotrophic setting, (1) to obtain more insights in the seasonal variabilityof coccolithophore assemblages and export production in response to hydrographicchanges, and (2) to quantify the coccolithophore-CaCO

3#ux. A comparison

of coccolithophore export production at 1000 m depth with coccolithophore #uxesat deeper deployed traps at station NABE-48 and with surface sediments is presentedin Ziveri et al. (2000), and those of the NABE-34 station will be publishedin a later stage. Additional data are already available from these sediment trapson biogenic #uxes (Honjo and Manganini, 1992, 1993), foraminifera (Wolfteich,1994) and diatoms (Takahashi, in Honjo and Manganini, 1993).

1878 A.T.C. Broerse et al. / Deep-Sea Research II 47 (2000) 1877}1905

Fig. 1. Map of surface water masses as described in literature for the northeast Atlantic Ocean. The twosediment trap locations are indicated as black dots (NABE-48: 483N 213W and NABE-34: 343N 213W).Surface water circulation: , after Sy (1988); , after Krauss and KaK se (1984); , after Worthington(1976); , after Olbers et al. (1982); , after Krauss (1988); , after Siedler et al. (1985); , after KaK seand Siedler (1982); , after McCartney and Talley (1982); SW, Subpolar Water; SF, Subarctic Front;NAC, North Atlantic Current; NATW, North Atlantic Transitional Water; AFZ, Azores Frontal Zone;(SC), (subtropical Convergence); AC, Azores Current; FJ, Frontal Jets; CC, Canary Current.

2. Materials and methods

2.1. Deployment and recovery of sediment traps

Sediment traps moorings were deployed at two sites in the NE Atlantic, at about343N 213W (NABE-34) and approximately 1445 km to the north at 483N 213W(NABE-48) (Fig. 1). These deployments were part of the JGOFS North AtlanticBloom Experiment (NABE) (Ducklow, 1989). Each mooring consisted of three auto-mated PARFLUX Mark 7G-13 time-series sediment traps (Honjo and Doherty,1988), attached to a mooring line at approximately 1, 2, and 3.7 km waterdepth. Thesea #oor depth at NABE-34 is approximately 5100 m and at NABE-48 4400 m.Deployment was from 3 April 1989 to 16 April 1990, with a hiatus from 26 Septemberto 16 October 1989 (Table 1). Individual sampling intervals for all deployments

A.T.C. Broerse et al. / Deep-Sea Research II 47 (2000) 1877}1905 1879

Table 1Summary table of North Atlantic Bloom Experiment mooring stations, showing location, bottom depth,trap depth, and deployment time

NABE-34 NABE-48

Phase 1 Phase 2 Phase 1 Phase 2

Latitude [email protected] N [email protected] N 47@. N [email protected] NLongitude [email protected] W [email protected] W [email protected] W [email protected] WBottom depth 5261 m 5083 m 4418 m 4451 mTrap depth!

1 km 1070 m 1248 m 1018 m 1202 m2 km 2067 m 1894 m 2018 m 2200 m3.7 km 4564 m 4391 m 3718 m 3749 m

!Phase 1: April 3, 1989 to September 26, 1989.Phase 2: October 16, 1989 to April 16, 1990.Hiatus: September 26, 1989 to October 16, 1989.

were generally 14 d, and each sediment trap was programmed to collect 26 samples(Table 2). Due to clogging of the shallow trap at 343N by an `argentine "sha, a part ofthe samples (3}13 and 27) is missing or unreliable (Table 2). A detailed description ofthe mooring design, traps deployment and analytical methods and results from allindividual samples is given in Honjo and Manganini (1992).

2.2. Coccolith analysis

For the coccolith analysis, aliquots of each sample were wet-split (aliquots rangedfrom 1/80 to 1/2000) using a rotary precision wet splitter (deviation between aliquots(3%). For splitting and sieving, rinse water was prepared from a 2% solution offormaldehyde in pre-"ltered sea water, which was bu!ered by adding sodium tetra-borate until pH 8.6 was obtained. The aliquots were then wet-sieved over a 32 lmmesh. Fecal pellets and aggregates'32 lm that remained on the sieve were disinteg-rated using a small soft brush and sieved again. The remaining '32 lm fraction(mainly hard-parts) was stored at 43C. The (32 lm fraction was "ltered overa 0.45 lm Millipore "lter of type HA (cellulose acetate; 47 mm diameter) and rinsedwith bu!ered distilled water to prevent crystallisation of salt. The "lters were dried at403C and stored in petri dishes.

For counting, a segment of the circular "lter was cut and laid in a drop ofimmersion oil on a microslide to obtain transparency, covered with a coverglass, and"xed with tape. For the quantitative analysis of coccoliths and coccospheres, we useda Olympus polarising light microscope at magni"cations of 1500]and 750], respec-tively. Coccoliths and coccospheres were counted along several parallel scans from theborder towards the center, according to the method described in Knappertsbusch andBrummer (1995). The total number of coccoliths counted in each sample rangedbetween 537 and 1385. Coccospheres were relatively rare compared to individualcoccoliths, and we identi"ed between 50 and 150 cells per sample. A JEOL JSM-840


Table 2Schedule showing mid-date, opening time, and coccolith recovery at sediment trap stations NABE-34 andNABE-48

Period Mid date Days open SamplesNABE-34

SamplesNABE-48

JD CD

1 96 04/06/89 5 p p2 105 04/15/89 14 p p3 119 04/29/89 14 m p4 133 05/13/89 17 m p5 148 05/29/89 14 m p6 164 06/13/89 14 m p7 178 06/27/89 14 m p8 192 07/11/89 14 n.p. p9 206 07/25/89 14 n.p. p

10 220 08/08/89 14 m p11 234 08/22/89 14 m p12 248 09/05/89 14 n.p. p13 262 09/19/89 14 m p14 279 10/06/89 20 (hiatus)15 296 10/23/89 14 p p16 310 11/06/89 14 p p17 324 11/20/89 14 p p18 338 12/04/89 14 p p19 352 12/18/89 14 p p20 1 01/01/90 14 p p21 15 01/15/90 14 p p22 29 01/29/90 14 p p23 43 02/12/90 14 p p24 57 02/26/90 14 p p25 71 03/12/90 14 p p26 85 03/26/90 14 p p27 99 04/09/90 14 n.p. pH

Hp"present, n.p."not present, m"mimumum #ux, JD"Julian date, CD"calender date.

scanning electron microscope (SEM) was used for taxonomic identi"cations of smallcoccoliths ((3 lm), except for those of Florisphaera profunda, whose #uxes weredetermined by light microscopy.

Coccolith and coccosphere counts were converted into #uxes (in number m~2d~1)by extrapolating the counted specimens to the entire e!ective "ltration area (1018 mm2)and to the total sample, considering duration days and trap aperture area.

Average daily #uxes and annual #uxes were normalized to 365 d, according to themethod described in Honjo and Manganini (1992). The average daily and annualcoccolithophore #uxes at NABE-34 are to be considered minimal, since we includedthe underestimated #uxes recorded during April 22 and October 16, 1989. However,(1) the strict similarity in composition, magnitude and trend of biogenic #uxes of theshallower trap with the deeper traps, (2) the high correlation (r2"0.80) between thetotal individual coccolith #ux and the total biogenic particle #ux at the shallow trap,


and (3) the low biogenic #uxes recorded at the deeper traps during the time intervalwhen the shallow trap was partly clogged, suggest low coccolith #uxes during thisperiod. Following these considerations, we expect only a slight underestimation of theannual coccolith #ux at NABE-34 obtained in the present study.

The Shannon diversity index (Shannon, 1949) is used to document diversity trendsthrough time, taking into account the relative proportion (of coccoliths) of species (cuto! level: 0.01%) in each sample.

In addition to the #uxes of coccolithophores, we determined #uxes of the calcareousdinophytes, mostly Thoracosphaera heimii (Tangen et al., 1982).

2.3. CaCO3

in the xne fraction

For the determination of carbonate weights in the "ne ((32 lm) fraction, we useda sample aliquot, equal to the one for coccolith #ux quanti"cation. The aliquot waswet-sieved over a 32 lm mesh, and the(32 lm fraction was "ltered over an acidi"edpre-weighted Isopore polycarbonate membrane "lter (0.4 lm pore size, 25 mm dia-meter). After drying at 403C, "lter and residue of the NABE-34 samples were leached in5 ml 10% HNO3~ solution for 12 h prior to the analysis, and those of NABE-48 in10 ml 0.1N HCl for 18 h. For logistic reasons Ca2` content of the samples was determinedby two methods. The samples from NABE-34 were analyzed by emission spectrometry(Perkin Elmer ICP/6500XR) and samples from NABE-48 by Flame Atomic AbsorptionSpectrometry (FAAS) using a Perkin Elmer model 2380. Overall accuracy amounted tobetter than 2% based on replicate analysis. Ca2` concentrations (in ppm) were thenconverted into CaCO

3#uxes (in mg m~2 d~1) by extrapolating the CaCO

3content of

each aliquot to the total sample, considering duration days and trap aperture area.These measured Ca}CaCO

3#ux in the "ne fraction was compared with the

summed CaCO3#ux of coccoliths, coccospheres and calcareous dinophytes. This

computed CaCO3#ux is obtained by multiplying the average mass of individual

species by their corresponding #ux. The coccolith and calcisphere masses used in thisstudy were determined by various scientists and based on species speci"c volumeestimates, derived from size measurements (Table 3). For Coccolithus pelagicus andCalcidiscus leptoporus, we carried out size measurements in the NABE-samples, andwe used the mass equation of Young and Ziveri (2000) to determine their coccolithmasses. For coccosphere-CaCO

3weight estimates, we used a mean number of

coccoliths per coccosphere (Table 3).

3. Taxonomic notes

Taxa were identi"ed by light microscopy (LM) and scanning electron microscopy(SEM). Small placoliths of E. huxleyi and gephyrocapsids (G. muellerae, G. ericsoniiand G. ornata) were counted by LM as one group, and so were Oolithotus fragilis andCalcidiscus leptoporus. We used SEM to determine the ratios of the group componentsby counting at least 50 specimens. Coccoliths of G. oceanica were identi"ed separatelyby LM.


Tab

le3

Ove

rvie

wof

cocc

olith

lengt

hsan

dm

asse

sofin

divi

dual

spec

ies

asde

term

ined

byva

riou

ssc

ientist

s,an

dth

em

ean

num

ber

ofco

ccol

iths

per

cocc

osph

ere

Spe

cies

Coc

colit

hC

occo

lith

Ref

eren

ceC

occo

liths

per

Ref

eren

cew

eigh

t(p

g)le

nght(l

m)

cocc

osph

ere

B.

bige

low

ii85

.52

7.1

Bea

ufo

rtan

dH

euss

ner

(199

9)C

.le

ptop

orus

(lar

ge)

110.

68!

You

ng

and

Ziv

eri(2

000)

31K

nap

pert

sbusc

h(1

993)

C.

lept

opor

us(sm

all)

74.1

7!Y

oung

and

Ziv

eri(2

000)

31K

nap

pert

sbusc

h(1

993)

C.

pela

gicu

s29

2.2

12.2

!Y

oung

and

Ziv

eri(2

000)

18M

cInty

rean

dB

eH(1

967)

D.

tubi

fera

1.5

3Y

oung

and

Ziv

eri(2

000)

E.

huxl

eyi

3.5

3.2

Knap

pert

sbusc

han

dB

rum

mer

(199

5)F

.pr

ofun

da1.

72.

5Y

oung

and

Ziv

eri(2

000)

76K

nap

pert

sbusc

h(1

993)

G.

ocea

nica

185

Knap

per

tsbu

sch

and

Bru

mm

er(1

995)

H.

cart

eri

142.

99.

11Bea

ufo

rtan

dH

euss

ner

(199

9)24

Knap

pert

sbusc

h(1

993)

O.

frag

ilis

96.8

8Y

oung

and

Ziv

eri(2

000)

P.

mul

tipo

ra65

.94

7.9

Bea

ufo

rtan

dH

euss

ner

(199

9)R

.cl

avig

era

67.5

10Y

oung

and

Ziv

eri(2

000)

smal

lge

phyr

ocap

sids

83.

9Y

oung

and

Ziv

eri(2

000)

S.pu

lchr

a16

6K

nap

per

tsbu

sch

and

Bru

mm

er(1

995)

smal

lSy

raco

spha

era

spp.

0.1

1.5

You

ng

and

Ziv

eri(2

000)

40K

nap

pert

sbusc

h(1

993)

U.

sibo

gae

sibo

gae

185

You

ng

and

Ziv

eri(2

000)

U.

sibo

gae

foli

osa

356

You

ng

and

Ziv

eri(2

000)

U.

tenu

is23

.86

5Bea

ufo

rtan

dH

euss

ner

(199

9)O

ther

s8

5J.

You

ng

(per

s.co

mm

.)20

"

T.he

imii

7560

15.5

Knap

pert

sbusc

han

dB

rum

mer

(199

5)

!Bas

edon

mea

sure

men

tsfrom

this

study

,w

ith

applic

atio

nof

equat

ion

inY

oun

g(this

issu

e).

"M

inim

um

valu

eus

edin

this

study

.


The size range of coccolith of C. leptoporus was large within a sample, and theaverage coccolith size varied considerably through time. In order to obtain reliableCaCO

3estimates of coccoliths of C. leptoporus we distinguished two types, based on

the number of elements of the distal shield, following the number of elements of TypeB and Type C of Kleijne (1993), and determined the average size of the two types ofcoccoliths (Table 3). The average size of distal shields with a maximum of 26 elementswas 7 lm, and distal shields with a minimum of 27 elements had an average size of8 lm. The ratio between the two forms has been determined by counting the numberof elements of at least 40 specimens per sample.

Except for Syracosphaera pulchra, the species of the genus Syracosphaera werecounted as one group (small Syracosphaera spp.), without distinguishing betweenindividual taxa, due to the di$culty in the identi"cation of coccospheres and indi-vidual coccoliths. Algirosphaera oryza and A. robusta were lumped as Algirosphaeraspp. for the same reason.

4. Hydrographic setting and particle 6uxes

4.1. Hydrographic setting

The circulation in the North Atlantic Ocean is characterized by the southward #owof the cold East Greenland and Labrador Currents and the northeastern #ow of theGulf Stream. The Labrador Current joins the warm, saline extension of the GulfStream o! Newfoundland to form the North Atlantic Current (NAC) (Krauss, 1986).The NAC current system is the main source of eddy kinetic energy for the northernNorth Atlantic (Krauss and KaK se, 1984). The northern branch of the NAC crosses theMid-Atlantic Ridge at about 523N into the northeast Atlantic and the southernbranch heads towards the Azores and continues to Madeira. A transitional zoneseparates the anticyclonic gyre of the subtropical waters to the south from the cyclonicgyre in the north (Krauss, 1986). The trap station at 483N (NABE-48) is located in thistransitional zone, south of the northern branch of the NAC (Fig. 1). The northernSubpolar Front (SF) at 513N shows a strong temperature gradient and separatesSubpolar Water (SW) from the NAC (Krauss and KaK se, 1984).

The relatively narrow southern branch of the NAC, known as the Azores Current(AC: KaK se and Siedler, 1982), originates in an area southeast of Newfoundland(Krauss et al., 1990). The Azores Frontal Zone (AFZ) separates at about 353Nnorthern waters with a seasonal thermocline formation from southwestern waterswith a permanent thermohaline mixed layer. Mesoscale eddy "elds are associated withthe frontal region of Azores Current (Wolfteich, 1994). The trap station at 343N(NABE-34) is located close to the path of the Azores Front (Fig. 1).

The monthly sea-surface temperatures of the studied areas for 1989 and 1990, basedon the Comprehensive Ocean-Atmosphere Data Set (COADS; Woodru! et al., 1987)are shown in Fig. 2. During the trap experiment local eddies developed near both trapsites, introducing colder, nutrient-rich water in the euphotic zone. Some of theseevents are not well re#ected in the monthly SST pro"le, due to the small scale andtemporal character of the eddy (Fig. 2).


Fig. 2. Monthly averaged Sea Surface Temperature (SST) in 1989 and 1990 (white dots) and long-termmonthly averaged wind speed from 1854}1979 (black squares) at NABE-34 and NABE-48. Data arederived from the COADS database of monthly composites within 0.53]0.53. At NABE-34, SSTs areaveraged for a region 33.5}34.53N 20.5}21.53W, and wind speeds for 33}353N 213W; at NABE-48, SST dataare from 483N 20.53W, and wind speeds are averaged for a region 47}493N 213W. Light shaded areasindicate the sampling period. Dark shaded areas indicate periods of mesoscale eddy development at thesediment trap sites (Wolfteich, 1994).

4.2. Particle yuxes

Honjo and Manganini (1993) identi"ed four major subsequent periods of mass#uxes during the experiment at both stations: bloom 1989, post-bloom, pre-bloom,and spring-bloom 1990 (Fig. 3). They de"ned the bloom episode as a rapid andcontinuous #ux increase, followed by a rapid and consistent decrease to the back-ground #ux. At both stations, a spring particle bloom consisted of two or threeoutstanding peaks separated by about a month.


Fig. 3. Annual variations in total mass and biogenic #uxes at NABE-34 and NABE-48 at 1 km depth (fromHonjo and Manganini, 1992). The "sh symbol indicates sample loss due clogging of the trap by a "sh(samples 3}13 and 27). H: Hiatus between "rst and second phase of trap deployment.

The annual mass #uxes during 1989/1990 at 1 km depth at NABE-34 and NABE-48were 19.4 and 20.7 g/m2, respectively, with average daily #uxes of 53.2 and56.7 mg/m2 (Honjo and Manganini, 1993). At both stations an episode of highparticle #uxes was recorded during the spring bloom (Fig. 3).

The annual opal #ux at 1 km depth was larger at NABE-48 than at the NABE-34(3.53 and 1.64 g/m2, respectively). The annual CaCO

3#uxes were 13.1 at NABE-34

and 11.0 g/m2 at NABE-48, representing, respectively, 67.6% and 55.3% of the mass#ux (Honjo and Manganini, 1993). All biogenic #uxes (opal, org. C, N, P and CaCO

3)

were highly correlated with the total mass #uxes at both stations (Table 4).Foraminifera #uxes were determined at NABE-34 in the 2 km trap, and at

NABE-48 a combination of samples from 2 km and 3.7 km depth was used (Wolfteich,1994). Foraminifera ('150 lm) reached highest #uxes during local bloom periods,and maximum #uxes at NABE-34 and NABE-48 were 1985 and 5100 speci-mens m~2d~1, respectively (Fig. 4; Wolfteich, 1994).

5. Results

5.1. Coccolithophore yuxes at stations NABE-34 and NABE-48

At both NABE-34 and NABE-48 stations, the presence of delicate coccolithophorespecies (e.g. holococcolithophores and Discosphaera tubifera), pteropods, and ascidianspicules in the sediment trap samples indicates that no signi"cant calcite dissolutionduring sedimentation to the shallow traps or during storage of the samples.

At station NABE-34 the coccolithophore #ux pattern showed a pronounced sea-sonal variation (Fig. 5). The coccosphere #ux increased in early January 1990 andreached a maximum value (43]106 coccospheres m~2 d~1) in late January (Fig. 5).The #ux of loose coccoliths increased more gradually, and a maximum of7.5]109 coccoliths m~2d~1 was recorded approximately 1.5 months later in March1990. The time-weighted mean coccolith and coccosphere #uxes were 1.12]109 and4.90]106 m~2 d~1, respectively. Calcareous dinophytes #uxes at NABE-34 station,


Tab

le4

Mat

rix

ofSpea

rman's

rank

corr

elat

ion

coe$

cien

tsbe

twee

nbio

genic#uxe

sat

NA

BE

-34

and

NA

BE-4

8

Cor

rela

tion

coe$

cien

t(r2)

343N

:n"

14;4

83N

:n"

26O

pal

(mg

m~

2d~

1)O

rg.

C,N

,P

(mg

m~

2d~

1)

Tota

lC

aCO

3(m

gm

~2d~

1)

CaC

O3(

32lm

(mg

m~

2d~

1)Fora

min

ifera

!

(no.m

~2d~

1)C

occo

lith

s(n

o.m

~2d~

1)C

occo

spher

es(n

o.m

~2d~

1)

Tota

lm

ass

343N

0.89

0.89

0.97

0.80

0.09

0.80

0.29

(mg

m~

2d~

1)483N

0.94

0.74

0.97

0.78

0.72

0.56

0.61

Opal

343N

*0.

780.

810.

580.

670.

520.

17(m

gm

~2d~

1)483N

*0.

670.

890.

780.

710.

560.

58O

rg.C

,N

,P

343N

**

0.77

0.52

0.32

0.71

0.12

(mg

m~

2d~

1)483N

**

0.60

0.67

0.42

0.66

0.68

Tota

lC

aCO

3343N

**

*0.

920.

480.

840.

40(m

gm

~2d~

1)483N

**

*0.

670.

780.

440.

49C

aCO

3(

32lm

343N

**

**

0.34

0.88

0.56

(mg

m~

2d~

1)483N

**

**

0.37

0.95

0.91

Fora

min

ifera

!343N

**

**

*0.

010.

24(n

o.m

~2d~

1)483N

**

**

*0.

200.

24C

occo

lith

s343N

**

**

**

0.25

(no.m

~2d~

1)483N

**

**

**

0.97

!Fora

min

ifera

'15

0lm

:N

ABE-3

4:#ux

esfrom

2000

mtr

apdep

th.N

ABE

-48:#ux

esfrom

2000

man

d37

00m

trap

sco

mbi

ned

.


Fig. 4. Annual variations in foraminiferal #uxes ('150 lm) at NABE-34 and NABE-48 (from Wolfteich,1994). Fluxes at NABE-34 are from 2 km depth, and at NABE-48 from 2 km (samples 1}17) and 3.7 kmdepth (samples 18}27). H: Hiatus between "rst and second phase of trap deployment.

Fig. 5. Annual variations in total coccolith (bars) and coccosphere (dots) #uxes at NABE-34 and NABE-48from the 1000 m trap. Note the di!erences in vertical scale. For other details, see Fig. 3.

mainly Thoracosphaera heimii, started to increase in November 1989 and reachedhighest values (838]103 calcispheres m~2 d~1) in January 1990 (Fig. 6).

At station NABE-48 coccolithophores showed an enhanced #ux during the spring1989, followed by low #uxes in summer and winter (Fig. 5). The #ux increased again inlate March 1990, and both coccoliths and coccospheres reached maximum#uxes simultaneously in early April (3.2]109 coccoliths m~2d~1 and 9.3]106coccospheres m~2d~1, respectively) (Fig. 5). The maximum coccolith and coccosphere#uxes were respectively 4 and 10 weeks delayed with respect to NABE-34. Thetime-weighted mean coccolith and coccosphere #uxes were 388]106 and8.45]105 m~2 d~1, respectively. Calcareous dinophytes, predominantly Thoracos-phaera heimii, showed increased #uxes during the spring of both 1989 and 1990(Fig. 6). Highest #uxes were recorded in May 1989 at 156]103 calcispheres m~2d~1.

5.2. Coccolithophore species composition at stations NABE-34 and NABE-48

At NABE-34 we identi"ed 43 heterococcolithophore and 8 holococcolithophoretaxa (Appendix A). Most species reached maximum coccolith and coccosphere #uxes


Fig. 6. Annual variations in calcareous dinophytes #uxes and their mass estimated CaCO3#uxes at

NABE-34 and NABE-48. For other details, see Fig. 3.

in late January or early March 1990 (Figs. 7 and 8). Emiliania huxleyi dominated thecoccosphere assemblage with an average 64%, and its contribution varied from 20%in autumn 1989 to '60% in spring 1990 (Fig. 9a). During November}December1989 coccospheres of Algirosphaera spp. and Gladiolithus yabellatus exhibited max-imum abundances. The loose coccolith assemblage was also dominated by E. huxleyi,mostly consisting of morphotype A (Young and Westbroek, 1991) (Fig. 9b). Its relativeabundance varied from 25% to 80%, with an average of 69%. Coccoliths of E. huxleyiwere most abundant during the summer and early autumn of 1989, and from January1990 until the end of the sampling period in April 1990. The second most abundantspecies in the coccolith assemblage, Florisphaera profunda, dominated during Novem-ber}December 1989, and contributed on average 12% (Fig. 9b). The Shannon diver-sity index (based on coccoliths) varied from 0.9 to 2.0 during the one-year experiment(Fig. 9c). Lowest diversity is exhibited during the spring periods of 1989 and 1990.

At NABE-48 we identi"ed 33 heterococcolithophore and 3 holococcolithophore taxa(Appendix A). Most species exhibited maximum #uxes in April 1990, except forCoccolithus pelagicus with maximum #uxes in June 1989 and holococcolithophores inAugust 1989 (Fig. 7). The dominant holococcolithophore species was Calyptrosphaeraoblonga, while C. pelagicus f. hyalinus was only rarely encountered. The main contribu-tors to the coccosphere assemblage were small gephyrocapsids (31%; mainly G. muel-lerae, few G. ericsonii), Syracosphaera spp. (27%), and E. huxleyi (18%). Coccospheres ofC. pelagicus dominated in June 1989 with 51% (Fig. 9a). The coccolith assemblagewas clearly dominated throughout the entire study period by E. huxleyi (rangingbetween 56 and 83% with on average 72%; Fig. 9b). Calcidiscus leptoporus was secondmost abundant in loose coccoliths with an average contribution of 7% (Fig. 9b).

The Shannon diversity index at NABE-48 was relatively constant and varied from0.7 to 1.5 (Fig. 9c). The decrease in diversity values during spring 1990 re#ects theincrease in relative abundance of E. huxleyi (Figs. 9b and c).

5.3. Coccolithophore contribution to the CaCO3yux at stations NABE-34 and NABE-48

At NABE-34 and NABE-48 the average "ne ((32 lm) fractions Ca}CaCO3

#ux were 22.2 and 12.5 mg m~2 d~1, respectively. The contribution of the "ne



b

Fig. 7. Annual variations in coccolith and coccosphere #uxes of key coccolithophore species at NABE-34and NABE-48. Coccolith #uxes are indicated by bars and coccosphere #uxes by dots. Note the variations invertical scale. For other details, see Fig. 3.

Fig. 8. Annual variations in coccolith (bars) and coccosphere #uxes (dots) of Algirosphaera spp. atNABE-34 and Oolithotus fragilis at NABE-48. For other details, see Fig. 3.

fraction Ca}CaCO3

to the total CaCO3

varied between 23% and 72% at NABE-34(Fig. 10a), and between 18% and 74% at NABE-48 (Fig. 10b), with average 62% and41%, respectively. During spring 1989, the contribution of "ne fraction CaCO

3to the

total CaCO3#ux at station NABE-48 was low, due to high foraminifera #uxes in

the fraction '150 lm (Fig. 4). During spring 1990, maximum contributions of the"ne fraction CaCO

3coincided with coccolithophore #ux maxima at both stations

(Figs. 5 and 10).The seasonal #ux pattern of the measured "ne fraction Ca}CaCO

3was

comparable to that of the summed estimated mass coccolith-, coccosphere- andcalcareous dinophyte-CaCO

3(Figs. 10c and d). However, at both stations the

summed estimated CaCO3

values were consistently lower than the measured"ne fraction Ca}CaCO

3values. The summed mass estimated CaCO

3#ux

of coccoliths, coccospheres and calcareous dinophytes amounted on average 55%at NABE-34 and 60% at NABE-48 to the measured "ne fraction Ca}CaCO

3#ux. Loose coccoliths were the main contributors to this mass estimated CaCO

3#ux

at both stations (with 78.8% at NABE-34 and 88.6% at NABE-48), while minorcontributions originated from coccospheres (6.6% and 7.2%, respectively) and cal-careous dinophytes (14.6% and 4.2%, respectively). During coccosphere #ux maxima,coccospheres made a signi"cant contribution to the estimated mass CaCO

3#ux (max.

18.2% at NABE-34 and 25.6% at NABE-48). At both stations the predominanceof coccolith-CaCO

3in the "ne fraction was also expressed in the high correlation

between the coccolith #ux and the "ne fraction Ca}CaCO3#ux (r2"0.88

at NABE-34 and r2"0.95 at NABE-48; Table 4). At NABE-34 station, the coccolith#ux was also highly correlated to the total CaCO

3#ux, while at NABE-48 the

correlation was lower due to high #uxes of foraminifera tests '150 lm (Table 4and Fig. 4).


Fig. 9. Relative abundances of the most common coccolithophore species at stations NABE-34 andNABE-48 for (a) coccospheres, and (b) coccoliths. (c) The Shannon diversity index is based on the relativeabundances of loose coccoliths. For other details, see Fig. 3.

6. Discussion

6.1. Coupling of biogenic yuxes with chlorophyll and nutrient concentrations

During spring 1990, biogenic #uxes at the northern station NABE-48 began toincrease in March 1990, approximately 56 d later than the onset of the spring particlebloom at NABE-34 (Honjo and Manganini, 1993). Unfortunately the sediment trapexperiment ended while #uxes at NABE-48 were still increasing, but during the springof 1989 the trap recorded maximum #uxes in May. The #ux pattern at NABE-34exhibited two peaks, the "rst in late January 1990, and the second in March 1990.Flux peaks in sediment traps are generally closely linked to the occurrence of


Fig. 10. Annual variations in "ne fraction ((32 lm) CaCO3

(our measurements) and its proportion(in %) to the total CaCO

3#ux as determined by Honjo and Manganini (1992) at (a) station NABE-34 and

(b) station NABE-48. A comparison of the annual variations in mass estimated CaCO3#uxes of coccoliths,

coccospheres and calcareous dinophytes, with the measured "ne fraction #uxes are shown in the lower"gures for (c) station NABE-34 and (d) station NABE-48. At NABE-34 "ne fraction CaCO

3#uxes were not

determined in samples 3}13 and 27, due to shortage or absence of material. H: Hiatus between "rst andsecond phase of trap deployment.

maximum chlorophyll concentrations in the surface waters with a time lag of 1}2months (Newton et al., 1994; Neuer et al., 1997). In the subtropical North AtlanticGyre, at approximately 303N 153W, monthly averaged chlorophyll values from 1979to 1985, derived from satellite imagery (CZCS), revealed an annual reoccurringmaximum of near surface chlorophyll concentrations (up to 0.2 mg/m3) in Decemberand January (Neuer et al., 1997). North of about 403N a rapid explosion of chloro-phyll pigments is detected each year in April or May (Ducklow and Harris, 1993). Atthis latitude, the spring bloom occurs in response to the seasonal development of thethermocline that shoals the mixed-layer depth after a period of nutrient replenishmentduring the winter mixing (Valiela, 1984). The #ux record at both NABE-34 andNABE-48 stations con"rms the apparent coupling of biogenic #uxes in sediment trapswith the reoccurring maxima in pigment concentrations in the surface waters duringlocal spring bloom development.

A close correlation between chlorophyll concentrations in surface waters andcoccolithophore #uxes in sediment traps has been recorded at station ESTOC, north


of Gran Canaria (293N 153W), at the eastern edge of the subtropical North AtlanticGyre (Fisher et al., 1996). Coccolithophore #uxes peaked in late January 1993, whilediatoms were of minor importance, corroborated by low biogenic silica #uxes (Fisheret al., 1996). Station NABE-34 revealed a comparable biogenic #ux record, withyear-round low silica #uxes, and high coccolithophore #uxes in January (Figs. 3}5).Therefore we assume that the #ux record at NABE-34 during the 1989}1990 deploy-ment re#ects a typical pattern for subtropical regions, although the timing of max-imum #uxes may vary interannually (Neuer et al., 1997).

At the subtropical NABE-34 station the average coccolith #ux was about a factorthree higher than at the temperate NABE-48 station (Fig. 5), whereas at the latterstation the biogenic silica #ux (mainly related to diatoms; Takahashi in Honjo andManganini, 1993) was signi"cantly higher (Fig. 3). At the latitude of NABE-48 stationdeep convection in late winter generally supplies 8 lmol/l of nitrate and 6 lmol/l ofsilicate in the mixed layer. At 343N, concentrations in the upper ocean do not exceed2}4 lmol/l of nitrate and 2 lmol/l of silicate (Glover and Brewer, 1988; Ducklow andHarris, 1993). Diatoms tend to dominate phytoplankton communities under high-nutrient conditions (Longhurst and Harrison, 1989), while most coccolithophorespecies are K-selected and dominate in more oligotrophic conditions (Brand, 1994).The initial high input of nitrate and silicate during winter mixing in the temperateAtlantic could enhance the rapid growing diatoms, while after silicate is depleted andnitrate is still present in low concentrations a community shift from diatoms tococcolithophores can be induced (Sieracki et al., 1993). In the more oligotrophic sitesin the subtropical Atlantic (e.g. stations NABE-34 and ESTOC) diatoms fail tobecome abundant, due to a low initial nutrient concentrations, and consequentlycoccolithophores tend to become the more numerous phytoplankton species.

The high average coccolith #ux (1120]106 m~2d~1) at station NABE-34 wascomparable to the #ux recorded at station Bermuda (313N 643W) in the westernsubtropical North Atlantic (1400]106 m~2d~1; Haidar et al., 2000). The averagecoccolith #uxes at the temperate station NABE-48 (388]106 m~2d~1) are slightlyhigher the #ux recorded in the tropical North Atlantic at 133N 543W(260]106 m~2 d~1 at 1000 m depth; Steinmetz, 1991). Lowest coccolith #uxes havebeen observed in the Arctic North Atlantic and they varied in 12}100]106 m~2d~1

(Samtleben and Bickert (1990) and recalculated from Andruleit (1997)).

6.2. Coupling of surface-water and coccolithophore yux events during 1989}1990

At station NABE-34 the "rst three samples (April 1989) record the end of the springbloom of 1989 (Honjo and Manganini, 1993). During this period coccolithophore#uxes were low and coccolithophore development had probably begun before thesediment trap experiment, as was shown by early coccosphere peak during the springbloom of 1990 (Fig. 5). This assumption is supported by Jochem and Zeitzschel (1993)and Passow and Peinert (1993), who found evidence for a transition phase betweena bloom and an oligotrophic system at the NABE-34 site during this period: theautotrophic biomass decreased from April 15}24, 1989, while nitrate concentrationswere (0.1 lmol/l down to 40}60 m. The nitrate maximum (up to 0.38 lmol/l)


occurred at 80}150 m depth, and a subsurface chlorophyll maximum had developed(Jochem and Zeitzschel, 1993). During the following period, the trap did not recordreliable #uxes, due to clogging of the trap. As already discussed, we assume that thecoccolithophore #ux in the summer months was low, corroborated by the lowbiogenic #uxes in the deep traps (Honjo and Manganini, 1993). In late summer, thenitrate depletion extended down to 100}120 m (Veldhuis et al., 1993) and we assumethat overall productivity was low due the continuing depletion of the euphotic zone.During winter, deep convection events in the North Atlantic are associated withenhanced wind stress, as is shown in the long-term "eld observations of windvelocities (COADS database; Fig. 2). A moderate increase in the monthly averagedwind speed is observed from September through February, followed by a decrease inspring (Fig. 2). The annual trend in the total coccosphere #ux exhibited a pronouncedincrease in January (Fig. 5). Results from a modeling study at station BATS in thewestern subtropical Atlantic show that winter surface irradiance is su$cient to initiatea bloom at Bermuda as soon as nutrients are entrained into the euphotic zone, anddoes not require the water column to restatify (Doney et al., 1996). Station NABE-34is located approximately at the same latitude, and we assume a comparable couplingbetween wind-induced nutrient entrainment and enhanced productivity in theeuphotic zone to be responsible for the increased coccosphere #ux in late January. Asshown by Wolfteich (1994), a mesoscale eddy might have attributed to the enhance-ment of the nutrients into the euphotic zone at the trap site NABE-34 during lateDecember 1989 and early January 1990.

At station NABE-48 the start of the 1989 spring bloom was triggered by a rapiddecrease in mixed-layer thickness after April 26 from approximately 130 to 20 m(Lochte et al., 1993). The onset of strati"cation coincided with the existence ofa mesoscale eddy "eld in the vicinity of the sediment trap station from April 27 to May3 (Marra and Ho, 1993; Robinson et al., 1993). The cold-core eddy was still present inthe trap area until at least May 24. Mixing processes, associated with the eddy, cane$ciently resupply nutrients into the surface waters after strati"cation (Garside andGarside, 1993). This was re#ected in an increase of the amount of chlorophyll a insidethe eddy core over the peripheral water masses by a factor two (Yoder et al., 1993).Diatom #uxes, derived from the sediment trap counts, reached a peak value in sampleNo. 4 from May 6 to 20 (Honjo and Manganini, 1993) and coincided with a signi"cantdecrease of dissolved silicate in surface waters, reaching a minimum concentration atMay 20, while nitrate concentrations were still declining (Sieracki et al., 1993).Maximum coccolithophore #uxes were recorded one sampling interval later, fromMay 20 to June 6, 1989. The succession in #ux peaks of diatoms to coccolithophoresin the trap samples also was exhibited in the phytoplankton assemblage in the surfacewater during the spring of 1989 (Sieracki et al., 1993; Lochte et al., 1993). During theR.V. Atlantis II cruises from May 18 to 31, 1989, the phytoplankton community shiftedin dominance from diatoms to small #agellates (presumed primarily prymnesiophytes;Sieracki et al., 1993). The increase in small #agellates started at May 25 and reachedmaximum concentrations on May 27. Data collected during Meteor cruises on April24 to May 31, 1989 revealed the same trend (Lochte et al., 1993). HPLC pigmentanalysis showed that fucoxanthin (indicative of diatoms) had highest concentrations


before May 15, while 19-hexaholyoxy-fucoxanthin (indicative of prymnesiophytes,here probably coccolithophores) increased after mid-May (Lochte et al., 1993). Thesuccession from a diatom- to a coccolithophore-dominated community appears to becaused by the complete depletion of dissolved silica whereas other nutrients are stillavailable in moderate concentrations (Sieracki et al., 1993). After the 1989 springbloom, coccolithophore #uxes decreased and remained low until the following spring(Fig. 5). During August 9}15, 1989, NABE-48 was situated within a large meander ofthe North Atlantic Current and a cyclonic eddy in the proximity of the trap developedfrom one of its smaller meanders (Wolfteich, 1994), resulting in a decrease of the SST(Fig. 2). The shoaling of the thermocline due to cyclonic circulation would displacenutrients towards the surface (Wolfteich, 1994). However, the coccolithophore #uxdata did not show a signi"cant response to the induced nutrient enrichment, incontrast to foraminifera, which exhibited increased #uxes during this period (Fig. 4).In late August 1989 the surface water conditions had returned to a typicaloligotrophic post bloom situation, with a sharp thermocline at 35 m depth andnutrient-depleted surface waters in which Cyanobacteria dominated the surface waterphytoplankton community (Veldhuis et al., 1993). The wind speed data at NABE-48show a pronounced seasonal variation, and the average wind velocity in winter is highcompared to that at station NABE-34 (Fig. 2). The strong wind-driven turbulence athigher latitudes will mix the coccolithophores below the photic zone. Under theseunfavourable conditions the coccolithophore #ux at the northern station appears tobe delayed until the water column stabilises, as is shown by the decreased wind stressand increased SST (see Fig. 2).

6.3. Seasonal changes in coccolithophore assemblages at stations NABE-34 andNABE-48

At the station NABE-34 the total number of coccolithophore species observed inthe sediment trap was higher than at NABE-48 (see Appendix A). The reduction inspecies richness from the subtropical to the temperate station supports the generalcoccolithophore biogeography in which the greatest species diversity is found in thetropical and subtropical zones and decreases towards the north (Winter et al., 1994).

The diversity index at NABE-48 is relatively constant and expresses the year-rounddominance of coccoliths of E. huxleyi. At the NABE-34 station, however, the #uctu-ations in species diversity are mainly induced by the increasing relative abundance ofcoccoliths of the deep-photic zone species Florisphaera profunda during the summerand autumn of 1989. The maximum relative abundance of F. profunda was reached inNovember and December, and corresponded with enhanced percentages of cocco-spheres of Algirosphaera spp. and Gladiolithus yabellatus. Florisphaera profunda andG. yabbelatus are exclusively abundant in low-latitude regions within the lower photiczone between 100}220 m (Okada and Honjo, 1973; Okada and McIntyre, 1977;Winter et al., 1994). The increasing proportion of deep living species is probablyrelated to the depletion of the upper photic zone during the summer, while nutrientsare still available in the lower part of the euphotic zone (Veldhuis et al., 1993). Themain picnocline was located at 150}200 m depth (Jochem and Zeitzschel, 1993) and


nutrients are continuously supplied to the lower part of the euphotic zone by di!usionthrough the picnocline, supporting the deep living community. A comparable sea-sonal succession, with F. profunda dominating in autumn and E. huxleyi during therest of the year, has been previously reported from the North Atlantic by Okada andMcIntyre (1979) in water samples (at 443N 413W) and by Haidar et al. (2000) in bothwater and sediment trap samples (at 313N 643W). At station NABE-48 #uxes ofdeep-dwelling coccolithophore species were low. However, Fig. 9a shows a slightincrease in the relative abundance of coccospheres of Algirosphaera spp. in Novemberand December 1989, during the same period as at the subtropical station. UnlikeF. profunda and G. yabbelatus, Algirosphaera spp. extends from subarctic to equatorialregions with increasing depth preference towards the lower latitudes (Honjo andOkada, 1974; Okada and McIntyre, 1979).

The cosmopolitan species E. huxleyi dominated the overall coccolith assemblage atboth sites. This is consistent with the distribution of this species in water samples fromthe western North Atlantic (Okada and McIntyre, 1979). The E. huxleyi #ux peak atNABE-34 during the end of January 1990 (6.0]109 coccoliths m~2d~1) is consider-ably higher than the maximum #ux recorded at NABE-48 (2.4]109 cocco-liths m~2d~1). Emiliania huxleyi becomes very abundant in a nutrient-richenvironment such as along the edges of the subtropical central gyres, in upwellingareas, and along the outer continental shelves (Brand, 1994). It occasionally reachbloom proportions ('1000 cells/ml) in the Atlantic, in between 503N and thesouthern coast of Iceland at about 633N, predominantly in the months of June andJuly (Tyrrell and Taylor, 1995). These blooms generally follow those of diatoms, andonly occur in highly strati"ed waters where the mixed-layer depth is usually inbetween 10 and 20 m, in conjunction with high surface irradiances (Holligan et al.,1993b; Tyrrell and Taylor, 1995; Nanninga and Tyrrell, 1996). A period of sunny andcalm weather is a requirement for bloom formation, and consequently blooms do notoccur every year. The station NABE-48 is located close to the proximity of thepotential bloom area of E. huxleyi, but its #ux pattern showed low values during thesummer months in which blooms usually occur. It is likely that the conditions duringthe trap deployment were unfavourable for a high E. huxleyi production.

Coccolithus pelagicus is a cold-water species and is dominant from sub-polar topolar waters in the northern Atlantic (McIntyre and BeH , 1967). At the subtropicalstation NABE-34, C. pelagicus was only rarely encountered. At station NABE-48, thecoccosphere #ux pattern clearly depicts an increase during June 1989, at the end of thespring-bloom (Fig. 7). A cyclonic cold-core eddy development from April 27 to May 3,1989 resulted in an increase in nutrient supply to the euphotic zone and a SST cooling(Wolfteich, 1994). Coccolithus pelagicus #uxes were higher at 2 and 3.7 km depth thanat 1 km depth (Ziveri et al., 2000). This anomaly could be caused by the lateral in#uxof C. pelagicus coccoliths and coccospheres during the settling process, in associationwith the cyclonic eddy. Cyclonic cold-core eddies in the northeastern Atlantic andtheir in#uence on biology has been investigated by Lochte and Pfannkuche (1987).The sampling stations of this former study were located in the proximity of theNABE-48 sediment trap site, at approximately 48}493N 213W, and were operativeduring May 1985. The investigation of Lochte and Pfannkuche (1987) showed that the


phytoplankton and bacteria had higher abundances, di!erent composition and di!er-ent distribution patterns within the eddy as compared to the ambient northeastAtlantic water. The anomaly of this colder water assemblage in#ux into the traps isalso proved by the coccolith composition of the surface sediments located underneaththe trap site. Coccolithus pelagicus was always less than 1% of the total coccolithassemblages in the sediment, while in the sediment trap assemblages average contribu-tions of 4% at 2 km and 3% at 3.7 km water depth were recorded (Ziveri et al., 2000).If we assume that interannual variability is negligible, we would have expected to "nda relative increase of this solution resistant species with respect to the more delicatecoccolith species in the coccolith assemblage of the surface sediment.

At NABE-48 holococcolithophore taxa (mainly Calyptrosphaera oblonga) revealedmaximum #uxes in August 1989, when low total coccolithophore #uxes were re-corded. Calyptrosphaera oblonga is known to have a world wide distribution and ismost frequently occurring holococcolithophore species in the eastern North Atlanticand the Mediterranean Sea (Kleijne, 1991). Maximum #uxes coincided with highsea-surface temperatures during the summer months, con"rming the observation ofJordan (1988) that they may be abundant during the summer in the northeastAtlantic.

Oolithotus fragilis is a species that prefers the mid- and lower-photic zone(80}200 m), and is found in temperate to tropical regions (Okada and McIntyre, 1977,1979; Okada and Honjo, 1973). Oolithotus fragilis was present at NABE-48 through-out the year and became one of the most abundant species in the coccolith assemblageduring spring 1990 (Figs. 8 and 9). At NABE-34 station, however, O. fragilis wasalmost absent. Its biogeographic boundaries in the surface sediments of the Atlanticfollow the 193C surface water isotherm in summer, and highest relative abundancesare recorded in subtropical and tropical regions (McIntyre and BeH , 1967). NABE-48station is located on the northern boundary of its distribution in the sediment.The #ux record in the sediment traps clearly deviates from the distribution ofcoccoliths of O. fragilis in the sediment. Coccoliths of O. fragilis are susceptible todissolution (Schneidermann, 1977), and it is likely that the species composition in thesurface sediment does not present a reliable re#ection of the distribution pattern of O.fragilis in the euphotic zone. This is corroborated by the pronounced decrease of theaverage coccolith #ux of O. fragilis with depth (15]106 m~2d~1 at 1 kmand 0.7]106 m~2d~1 at 3.7 km (Ziveri et al., 2000)) and by the reduction in itscontribution to the total coccolith assemblage (3.8% at 1 km and (1% at 3.7 km(Ziveri et al., 2000)).

Calcidiscus leptoporus it is one of the most eurythermal coccolithophore species,ranging from equatorial to polar waters, although it never constitutes a large part ofthe living #ora (McIntyre and BeH , 1967). At both trap sites coccolith #uxesof C. leptoporus were present throughout the year and it showed only moderatelyincreased #uxes during local spring periods (Fig. 7). Although the coccolith #uxmagnitude of C. leptoporus were comparable at both trap sites, its relative abundanceswere signi"cantly higher at the temperate station, where it was the second mostabundant species in the coccolith assemblage (7% at NABE-48 and 2% at NABE-34).The relative increase of C. leptoporus has been assigned as an important characteristic


in de"ning the transitional zone from warm to colder #oras, together with theappearance of C. pelagicus (McIntyre and BeH , 1967).

Syracosphaera pulchra thrives in a wide temperature range (Okada and McIntyre,1979). The magnitude of coccolith #uxes of S. pulchra were comparable at both sites,and maximum #uxes were recorded during local spring bloom periods. At NABE-48station S. pulchra exhibited maximum #uxes during the 1990 spring-bloom, while nosigni"cant peak was recorded during the 1989 spring-bloom. A sediment trap experi-ment in the vicinity of NABE-48, shortly after the end of our experiment, revealeda short and rapid #ux increase of S. pulchra in early June 1990 (Knappertsbusch andBrummer, 1995). Another sediment trap experiment at 473N 193W recorded twoseparate coccolith #ux peaks, the "rst in late April 1990 and the second in late June1990 (Newton et al., 1994). Unfortunately our experiment ended in April, but weassume from the combined sediment trap records that during the spring bloom of1990 a coccolithophore succession from E. huxleyi to S. pulchra had occurred.

6.4. Coccolithophore-CaCO3yuxes at stations NABE-34 and NABE-48

Carbonate export production in the open ocean generally ranges from about2}5 g m~2 yr~1 in the central ocean gyres to as much as 30}40 g m~2 yr~1 in theeastern upwelling zones (Milliman, 1993). The annual CaCO

3#ux in the North Atlantic

at the subtropical and the temperate station (13.1 and 11.1 g m~2 yr~1, respectively), asdetermined by Honjo and Manganini (1993), falls well in between these extremes.Carbonate export #uxes mainly consists of foraminifera and coccolithophores, and toa lesser extend of Thoracosphaera spp. and pteropods (Takahashi, 1994).

Previous studies on the determination of coccolithophore-CaCO3

masses werebased on geometric measurements of coccoliths of various species (Samtleben andBickert, 1990; Knappertsbusch and Brummer, 1995; Beaufort and Heussner, 1999;Young and Ziveri, 2000). Young and Ziveri (2000) have calculated an error incoccolith mass estimates of around 50%, in addition to errors due to incompletebreakdown of pellets and aggregates. Furthermore, the exclusion of coccolithfragments from the countings will underestimate this value. In the present studywe used the "ne fraction ((32 lm) Ca}CaCO

3#ux as an additional estimation

of coccolithophore-CaCO3#ux. This method avoids the errors derived from

coccolithophore #ux and coccolith mass estimations. However, in the "ne fractionother CaCO

3particles also are present, including detrital carbonate and biogenic

CaCO3

(juvenile tests and fragments of foraminifera, Thoracosphaera spp., pteropodfragments, ascidian spicules, and larval shells of mollusks).

Microscopic observations revealed that individual coccoliths were the main con-tributors to the "ne fraction carbonate at NABE-34 and NABE-48, as also shown bythe high correlation coe$cient between the number of individual coccoliths and thetotal "ne fraction Ca}CaCO

3#ux (Table 4). In this study, CaCO

3from small

foraminifera, pteropods, and non-biogenic sources in the "ne fraction was minimal,and average CaCO

3#uxes from calcareous dinophytes were low (0.9 g m~2 yr~1 at

NABE-34 and 0.1 g m~2 yr~1 at NABE-48). The computed annual coccolithophore(coccoliths and coccospheres) #ux, based on estimates of individual coccolith masses,


at NABE-34 and NABE-48 was 5.1 and 2.7 g m~2 yr~1, respectively, representing39% and 24% to the total CaCO

3#ux. The annual "ne fraction Ca}CaCO

3#ux was

signi"cantly larger, with 8.1 g m~2 yr~1 at NABE-34 and 4.6 g m~2 yr~1 at NABE-48, providing an average contribution to the total #ux of 62 and 41%, respectively. Atboth stations, the correlation between the summed mass estimated CaCO

3#uxes

(including coccoliths, coccospheres and calcareous dinophytes) and "ne fractionCaCO

3#ux is high (Spearman correlation by ranks: r2"0.97 at NABE-34 and

r2"0.86 at NABE-48), which indicates that no major carbonate sources in the "nefraction are excluded in our calculations. Therefore we assume that the discrepancybetween coccolithophore-CaCO

3estimates resulting from the two methods is prim-

arily related to an underestimation of the coccolith mass estimated #ux of individualspecies and cannot be only related to the presence of other CaCO

3particles.

7. Summary and conclusions

1. Distinct seasonal changes in coccolithophore #uxes occurred during the1989}1990 sediment trap deployments at the subtropical station NABE-34 andthe temperate station NABE-48, where 92% and 70%, respectively, of the coccolithswere trapped during local particle spring bloom periods. During spring 1990, cocco-sphere #uxes started to increase at NABE-34 in January, and at NABE-48 approxim-ately two months later, in March. At both stations, the variations in coccolithophore#uxes re#ect the response of coccolithophores to the seasonal changes in upper watercolumn processes.

2. A comparison of stations NABE-34 and NABE-48 show higher average coccolith#uxes (1,120]106 coccoliths m~2d~1 vs. 388]106 coccoliths m~2d~1) and higherspecies richness (51 vs. 36 species) at the subtropical station NABE-34. This preferenceof coccolithophores for more oligotrophic conditions is probably related to a reducedcompetition with diatoms in the subtropical Atlantic.

3. At both the subtropical and the temperate stations, Emiliania huxleyi was themain contributor to the numerical annual coccolith #ux (with 69% and 72%,respectively). At NABE-34, the relative abundance of E. huxleyi was highest duringspring 1990. Coccoliths of Florisphaera profunda increased through summer andautumn, probably related to the progressive depletion of nutrients in the euphoticzone. At NABE-48, coccoliths of E. huxleyi dominated throughout the year, followedby C. leptoporus and small gephyrocapsids.

4. The "ne fraction ((32 lm) CaCO3#ux at NABE-34 was 8.1 g m~2 yr~1 and at

NABE-48 4.6 g m~2 yr~1, representing 62% and 41% of the total CaCO3#ux. The

mass estimated CaCO3#ux of coccoliths and coccospheres was 5.1 g m~2 yr~1 at

NABE-34 and 2.7 g m~2 yr~1 at NABE-48, and that of calcareous dinophytes0.9 g m~2 yr~1 and 0.1 g m~2 yr~1, respectively. The summed mass estimatedCaCO

3#ux of loose coccoliths, coccospheres and calcareous dinophytes was highly

correlated to the measured "ne fraction Ca}CaCO3#ux at both stations (r2"0.97

at NABE-34 and r2"0.86 at NABE-48). However, the summed mass estimatedvalues were consistently lower, and contributed only 55% at NABE-34 and 60% at


NABE-48 to the measured "ne fraction Ca}CaCO3#ux, although microscopic

observations revealed no major "ne fraction CaCO3

sources besides coccolithophoresand calcareous dinophytes. The two methods of determining coccolithophore-CaCO

3#uxes induce errors. We attribute the present discrepancy in the results of the twomethods primarily to an underestimation of the mass estimated CaCO

3#ux.

Acknowledgements

This study was part of the `Global Emiliania Modeling Initiativea and was sup-ported by the Netherlands Geosciences Foundation (GOA) with "nancial aid fromthe Netherlands Organisation for Scienti"c Research (NWO). We are grateful toSteven Manganini from the Woods Hole Oceanographic Institution for providing thesediment trap aliquots. We thank Saskia Kars for operating the SEM and makingphotographs, Peter Willekes for assisting with the sample preparation, Geert-JanBrummer and John Kist for the calcium analysis. We also thank H. Andruleit,K-H. Baumann, and M. Knappertsbusch for their helpful comments on the manu-script. This paper is publication no. 981103 of the Netherlands Research School ofSedimentary Geology (NSG), Vrije Universiteit, Amsterdam.

Appendix A

Coccolithophore species composition during the 1989}1990 JGOFS sediment trapexperiment at stations NABE-34 and NABE-48.

NABE34 48

HeterococcolithophoresAcanthoica sp. Lohmann (1903), emend. Schiller (1913) and Kleijne (1992) *Algirosphaera oryza Schlauder (1945) * *Algirosphaera robusta (Lohmann 1902) Norris (1984) *Alisphaera unicornis Okada and McIntyre (1977) * *Braarudosphaera bigelowii (Gran and Braarud, 1935) De#andre (1947) * *Calcidiscus leptoporus (Murray and Blackman, 1898) Loeblich and Tappan (1978) * *Calciosolenia murrayi Gran (1912) * *Ceratolithus cristatus Kamptner (1950) var. cristatus * *Coccolithus pelagicus f. hyalinus (Gaarder and Markali, 1956) Kleijne (1991) * *Coccolithus pelagicus (Wallich, 1877) Schiller (1930) f. pelagicus * *Coronosphaera binodata (Kamptner, 1927) Gaarder, in Gaarder and Heimdal (1977) * *Coronosphaera mediterranea (Lohmann, 1902) Gaarder in Gaarder and Heimdal (1977) * *Discosphaera tubifera (Murray and Blackman, 1898) Ostenfeld (1900) * *Emiliania huxleyi (Lohmann, 1902) Hay and Mohler, in Hay et al. (1967) var. huxleyi * *Florisphaera profunda Okada and Honjo (1973) var. profunda Okada and McIntyre (1977) * *Gephyrocapsa ericsonii McIntyre and BeH (1967) * *Gephyrocapsa muellerae BreH heH ret (1978) * *Gephyrocapsa oceanica Kamptner (1943) * *


Gephyrocapsa ornata Heimdal (1973) *Gladiolithus yabellatus (Halldal and Markali, 1955) Jordan et Chamberlain (1993) *Hayaster perplexus (Bramlette and Riedel, 1954) Bukry (1973) *Helicosphaera carteri (Wallich, 1877) Kamptner (1954) var. carteri * *Helicosphaera pavimentum Okada and McIntyre (1977) *Neosphaera coccolithomorpha Lecal-Schlauder (1950) *Oolithotus fragilis (Lohmann, 1912) Martini and MuK ller (1972) var. fragilis * *Papposphaera sp. Tangen (1972) *Pontosphaera discopora Schiller (1925) * *Pontosphaera japonica (Takayama, 1967) Nishida (1971) *Pontosphaera syracusana Lohmann (1902) * *Reticulofenestra parvula (Okada and McIntyre, 1977) Biekart (1989) var. parvula *Reticulofenestra sessilis (Lohmann, 1912) Jordan and Young (1990) *Rhabdosphaera clavigera Murray and Blackman (1898) * *Scyphosphaera apsteinii Lohmann (1902) f. apsteinii * *Syracosphaera borealis Okada and McIntyre (1977) * *Syracosphaera corolla Lecal (1966) *Syracosphaera molischii Schiller (1925) * *Syracosphaera nana (Kamptner, 1941) Okada and McIntyre (1977) * *Syracosphaera nodosa Kamptner (1941) *Syracosphaera pulchra Lohmann (1902) * *Syracosphaera variabilis (Halldal and Markali) Okada and McIntyre (1977) * *Umbellosphaera irregularis Paasche (1950) * *Umbellosphaera tenuis (Kamptner, 1937) Paasche (1955) * *Umbilicosphaera hulburtiana Gaarder (1970) *Umbilicosphaera sibogae var. foliosa (Kamptner, 1963) Okada and McIntyre (1977) * *Umbilicosphaera sibogae (Weber-van Bosse, 1901) Gaarder (1970) var. sibogae * *

HolococcolithophoresCalyptrosphaera oblonga Lohmann (1902) * *Corisphaera gracilis Kamptner (1937) *Daktylethra pirus (Kamptner, 1937) Norris (1985) *Gliscolithus amitakarense Norris (1985) *Helladosphaera sp. Kamptner (1937) *Homozygosphaera vercellii Borsetti and Cati (1979) *Periphyllophora mirabilis (Schiller, 1925) Kamptner (1937) * *Poricalyptra aurisinae Kamptner *Zygosphaera marsilii (Borsetti and Cati, 1976) Heimdal (1982) *

References

Andruleit, H., 1997. Coccolithophore #uxes in the Norwegian-Greenland Sea: seasonality and assemblagealterations. Marine Micropaleontology 31, 45}64.

Berger, W.H., 1976. Biogeneous deep sea sediments: production, preservation and interpretation..In: Riley, J.P., Chester, R. (Eds.), Chemical Oceanography, Vol. 5. Academic Press, New York,pp. 265}388.

Beaufort, L., Heussner, S., 1999. Coccolithophorids on the continental slope of the Bay of Biscay 1.Production, transport and contribution to mass #uxes. Deep-Sea Research II 46, 2147}2174.

van Bleijswijk, J.D.L., 1996. Ecophysiology of the calcifying marine alga Emiliania huxleyi. Ph.D. Disserta-tion, Rijks Universteit Groningen, 71 pp.

Brand, L.E., 1994. Physiological ecology of marine coccolithophores. In: Winter, A., Siesser, W.G. (Eds.),Coccolithophores.. Cambridge University Press, Cambridge, pp. 39}49.


Brown, C.W., Yoder, J.A., 1994. Distribution pattern of coccolithophorid blooms in the Western NorthAtlantic Ocean. Continental Shelf Research 14, 175}197.

Doney, S.C., Glover, D.M., Najjar, R.G., 1996. A new coupled, one-dimensional biological-physical modelfor the upper ocean: applications to JGOFS Bermuda Atlantic Time-series Study (BATS) site. Deep-SeaResearch II 43 (2/3), 591}624.

Ducklow, H.W., 1989. Joint global ocean #ux study: the North Atlantic bloom experiment. OceanographyMagazine 2, 4}8.

Ducklow, H.W., Harris, R.P., 1993. Introduction to the JGOFS North Atlantic bloom experiment.Deep-Sea Research II 40 (1/2), 1}8.

Fisher, G., Neuer, S., Krause, G., Wefer, G., 1996. Short-term sedimentation pulses recorded witha #uorescence sensor and sediment traps in 900 m water depth in the Canary basin. Limnology andOceanography 41, 1354}1359.

Gaarder, K.R., 1971. Comments on the distribution of coccolithophorids in the ocean. In: Funnell, B.M.,Riedel, W.R. (Eds.), The Micropaleontology of Oceans. Cambridge Univ. Press, Cambridge,pp. 97}103.

Garside, C., Garside, J.C., 1993. The `f-ratioa on 203W during the North Atlantic Bloom Experiment.Deep-Sea Research II 40 (1/2), 75}90.

Glover, D.M., Brewer, P.G., 1988. Estimates of wintertime mixed layer concentrations in the NorthAtlantic. Deep Sea Research 35 (9), 1525}1546.

Haidar, A.T., Thierstein, H.R., Deuser, W.G., 2000. Calcareous phytoplankton standing stocks, #uxes andaccumulation in Holocene sediments o! Bermuda (N. Atlantic). Deep-Sea Research II 47, 1907}1938.

Holligan, P.M., Fernandez, E., Aiken, J., Balch, W.M., Boyd, P., Burkill, P.H., Finch, M., Groom, S.B.,Malin, G., Muller, K., Purdie, D.A., Robinson, C., Trees, C.C., Turner, S.M., van der Wal, P., 1993a.A biogeochemical study of the coccolithophore Emiliania huxleyi in the North Atlantic. GlobalBiogeochemical Cycles 7, 879}900.

Holligan, P.M., Groom, S.B., Harbour, D.S., 1993b. What controls the distribution of the coccolithophore,Emiliania huxleyi, in the North Sea. Fisheries Oceanography 2, 175}183.

Honjo, S., Doherty, K.W., 1988. Large aperture time-series sediment traps; design objectives, constructionand application. Deep-Sea Research 35, 133}149.

Honjo, S., Manganini, S.J., 1992. Biogenic particle export to the interior of the North Atlantic Ocean at the343 and 483N stations, 1989/1990: method and analytical data compilation. WHOI Technical Report,WHOI-92-15, 74 pp.

Honjo, S., Manganini, S.J., 1993. Annual biogenic particle #uxes to the interior of North Atlantic Ocean;Studied at 343N 213W and 483N 213W. Deep-Sea Research II 40 (1/2), 587}607.

Honjo, S., Okada, H., 1974. Community structure of coccolithophores in the photic layer of the mid-Paci"c.Micropaleontology 29, 209}230.

Jochem, F.J., Zeitzschel, B., 1993. Productivity regime and phytoplankton size structure in the tropical andsubtropical North Atlantic in spring 1989. Deep-Sea Research II 40, 495}519.

Jordan, R.W., 1988. Cocccolithophorid communities in the northeast Atlantic. Unpublished Ph.D. Thesis,University of Surrey, 350 pp.

KaK se, R.H., Siedler, G., 1982. Meandering of the subtropical front southeast of the Azores. Nature 300,245}246.

Kleijne, A., 1991. Holococcolithophorids from the Indian Ocean, Red Sea, Mediterranean Sea and NorthAtlantic Ocean. Marine Micropaleontology 17, 1}76.

Kleijne, A., 1993. Morphology, taxonomy, and distribution of extant Coccolithophorids. Ph.D. Disserta-tion, Free Univiversity, Amsterdam, 321 pp.

Knappertsbusch, M., 1993. Geographic distribution of living and Holocene coccolithophores in theMediterranean Sea. Marine Micropaleontology 21, 219}247.

Knappertsbusch, M., Brummer, G.-J.A., 1995. A sediment trap investigation of sinking coccolithophorids inthe North Atlantic. Deep-Sea Research 7, 1083}1109.

Krauss, W., 1986. The North Atlantic Current. Journal of Geophysical Research 91 (c4), 5061}5074.Krauss, W., DoK scher, R., Lehmann, A., Vieho!, T., 1990. On eddy scales in the eastern and northern North

Atlantic Ocean as a function of latitude. Journal of Geophysical Research 95, 18 049}18 056.


Krauss, W., KaK se, R.H., 1984. Mean circulation and eddy kinetic energy in the eastern North Atlantic.Journal of Geophysical Research 89, 3407}3415.

Lochte, K., Pfannkuche, O., 1987. Cyclonic cold-core eddy in the eastern North Atlantic. II. Nutients,phytoplankton and bacterioplankton. Marine Ecology Progress Series 39, 153}164.

Lochte, K., Ducklow, H.W., Fasham, M.J.R., Stienen, C., 1993. Plankton succession and carbon cycling at473N 203W during JGOFS North Atlantic Bloom Experiment. Deep-Sea Research II 40, 91}114.

Longhurst, A.R., Harrison, W.G., 1989. The biological pump: pro"les of plankton production andconsumption in the upper ocean. Progress in Oceanography 22, 47}123.

Marra, J., Ho, C., 1993. Initiation of the spring bloom in the northeast Atlantic (473N, 203W): a numericalsimulation. Deep-Sea Research II 40, 55}73.

Olbers, D.J., Willebrand, J., Wensel, M., 1982. Ocean Modelling 46 (unpublished manuscript).Passow, U., Peinert, R., 1993. The role of plankton in particle #ux: two case studies from the northeast

Atlantic. Deep-Sea Research II 40, 573}585.Robinson, A.R., McGillicuddy, D.J., Calman, J., Ducklow, H.W., Fasham, M.J.R., Hoge, F.E., Leslie, W.G.,

McCarthy, J.J., Podewski, S., Porter, D.L., Saure, G., Yoder, J.A., 1993. Mesoscale and upper oceanvariabilities during the 1989 JGOFS bloom study. Deep-Sea Research II 40, 9}35.

Samtleben, C., Bickert, T., 1990. Coccoliths in sediment traps from the Norwegian Sea. Marine Micro-paleontology 16, 39}64.

Schneidermann, N., 1977. Selective dissolution of recent coccoliths in the Atlantic Ocean.. In: Ramsay,A.T.S. (Ed.), Oceanic Micropalaeontology, Vols. 1 and 2. Academic Press, London, pp. 1009}1053.

Shannon, C.E., 1949. The mathematical theory of environments.. In: Shannon, C.E., Weaver, W. (Eds.), TheMathematical Theory of Communication.. University of Illinois Press, Urbana, pp. 1}93.

Siedler, G., Zenk, W., Emery, W.J., 1985. Strong current events related to a subtropical front in thenortheast Atlantic. Journal of Physical Oceanography 15, 885}897.

Sieracki, M.E., Verity, P.G., Stoecker, D.K., 1993. Plankton community response to sequential silicate andnitrate depletion during the 1989 North Atlantic spring bloom. Deep-Sea Research II 40 (1/2), 213}226.

Steinmetz, J.C., 1991. Calcareous Nannoplankton Biocoenosis: Sediment Trap Studies in the EquatorialAtlantic, Central Paci"c, and Panama Basin. Ocean Biocoenosis Series No. 1, Woods Hole Oceano-graphic Institution, 85 pp.

Sy, A., 1988. Investigation of large scale circulation patterns in the Central North Atlantic: the NorthAtlantic Current, the Azores Current and the Mediterranean water plume in the area of the MidAtlantic Ridge.. Deep-Sea Research 25 (3), 383}413.

Takahashi, K., 1994. Coccolithophorid biocoenosis: production and #uxes to the deep sea.. In: Green, J.C.,Leadbeater, B.S.C. (Eds.), The Haptophyte Algae. Systematic Association Special Vol. 51. ClarendonPress, Oxford, pp. 335}350.

Tangen, K., Brand, L., Blackwelder, P.L., Guillard, R.R.L., 1982. Thoracosphaera heimii (Lohmann)Kamptner is a dinophyte: observations on its morphology and life cycle. Marine Micropaleontology 7,193}212.

Tyrrell, T., Taylor, A.H., 1995. A modelling study of Emiliania huxleyi in the NE Atlantic. Journal of MarineSystems 9, 83}112.

Valiela, I., 1984. Marine Ecological Progresses. Springer, New York, 546 pp.Veldhuis, M.W.J., Kraay, G.W., Gieskes, W.W.C., 1993. Growth and #uorescence characteristics of

ultraplankton on a north-south transect in the eastern North Atlantic. Deep-Sea Research II 40 (1/2),609}626.

Westbroek, P., Brown, C.W., Van Bleijswijk, J., Brownlee, C., Brummer, G.J., Conte, M., Egge, J.,FernaH ndez, E., Jordan, R., Knappertsbusch, M., Stefels, J., Veldhuis, M., Van der Wal, P., Young, J.,1993. A model system approach to biological climate forcing: the example of Emiliania huxleyi. Globaland Planetary Change 8, 27}46.

Winter, A., Jordan, R., Roth, P., 1994. Biogeography of living coccolithophores in ocean waters. In: Winter,A., Siesser, W.G. (Eds.), Coccolithophores. Cambridge University Press, Cambridge, UK, pp. 161}177.

Wolfteich, C.M., 1994. Satellite-derived sea surface temperature, mesoscale variability, and foraminiferalproduction in the North Atlantic. M.Sc., Massachusetts Institute of Technology/Woodshole Oceano-graphic Institution, 80 pp.


Woodru!, S.D., Slutz, R.J., Jenne, R.L., Steurer, P.M., 1987. A comprehensive ocean-atmosphere data set.Bulletin American Meteorological Society 68, 1239}1249.

Worthington, C.V., 1976. On the North Atlantic circulation. John Hopkins University OceanographyStudies 6, 110.

Yoder, J.A., Aiken, J., Swift, R.N., Hoge, F.E., Stegmann, P.M., 1993. Spatial variability in near-surfacechlorophyll a #uorescence measured by Airborne Oceanographic Lidar (AOL). Deep-Sea Research II40 (1/2), 37}54.

Young, J.R., Westbroek, P., 1991. Genotypic variation in the coccolithophorid species Emiliania huxleyi.Marine Micropaleontology 18, 5}23.

Young, J.R., Ziveri, P., 2000. Calculation of coccolith volume and its use in calibration of carbonate #uxestimates. Deep-Sea Research II 47, 1679}1700.

Ziveri, P., Broerse, A.T.C., van Hinte, J.E., Westbroek, P., Honjo, S., 2000. The fate of coccoliths at 483N213W, northeastern Atlantic. Deep-Sea Research II 47, 1853}1875.


Coccolithophore export production, species composition, and coccolith-CaCO3 fluxes in the NE Atlantic ( and )

Documents