Download Accumulation of Th, Pb, U, and Ra in marine phytoplankton and its

32(l), 1987,l3l-142
Limnol. Oceanogr.,
Inc.
@ 1987,by the Amcrican SocietyofLimnology and OccanoS;raphy,
Accumulation of Th, Pb, U, and Ra in marine phytoplankton and
its geochemicalsignificancet
N. S. Frsfter2and J.-L. TeyssiE
International I"aboratory of Marine Radioactivity, Mus6e Oceanographique, MC 98000 Monaco
S. Krishnaswami and M. Baskaran
Physical ResearchLaboratory, Ahmedabad 380009 India
Abstract
The bioaccumulation of U, Th, Ra, and Pb in four diverse nanoplanktoais algal species and a
picoplanktonic blue-green alga was determined with radiotracers. Among the nanoplankton, diflerences of l-2 orderi of magnitude in volume/volume concentration factors (V'CFs) were observed
for a given nuclide, but larger diferences were observed among the four nuclides, with VCF values
of Th > Pb > Ra = U. The picoplanlcon cells, with greater surface: volume ratios, had significantly
higher VCF values. The mean VCF values in the nanoplankton of Th and Pb were 1.5 x 105and
3.[xlOaintheliehtand2.8xlOsandT.3xl0linthedark.TheVCFsofThandPbinthe
22tTh and
picoplankton were both about 2 x 106, irrespective of light. Retention half-times of
iropf io fecal pellets of Artemia salina, fed radiolabeled diatoms, were 20-50 d, but >120 d for
22tTh at 4"C. ihe results suggestthat sinking plankton and their debris could a@ount for most of
the natural series radionuclides sedimenting out ofoceanic surfacc waters.
It is well established that several natural
radionuclides, 234Th,228Th,2r0Pb,and 2r0Po,
have very short residencetimes in oceanic
surfacewaters (Cochran |982;Broecker and
Peng 1982).The rapid removal of thesenuclides has led to the suggestionthat they are
transported from surface to deep waters
through biologically mediated processes
(Rama et al. 1961; Bhat et al. 1969;Broecker et al. t973;Matsumoto 1975).The recent
observations that the fluctuations in the settling fluxes of 228Thin the deep waters of
the SargassoSea (Bacon et al. 1985) are
strongly correlated with the primary productivity in the euphotic zone and that the
scavenging rate of 234This proportional to
primary production in the California Current (Coaleand Bruland 1985)support these
earlier suggestions. However, direct mea"net
surementsof Th isotopesand 2roPbin
plankton" yielded low concentrations of
these nuclides in this material, leading to
the suggestion that biological uptake may
be unimportant in the transport of these
nuclides from surface to deep waters of the
'Supported in part by IAEA Contract 3209 to the
Physical Researchl,aboratory, Ahmedabad, India.
2 Presentaddress:OceanographicSciencesDivision,
Brookhaven National Laboratory, Upton, New York
tt973.
open ocean(Krishnaswami et al. 1976;Nozaki and Tsunogai 1976; Turekian 1977).
We have examined this problem in more
detail, since critical evaluation of the role
of marine organisms in mediating the distribution of these nuclides in surfacewaters
requires knowledge of their concentration
in the primary producers. The net planllon
used for calculating the export fluxes ofthese
nuclides out of the euphotic zone would be
dominated by zooplanklon and larger phytoplankton and hence would not be representative of the preponderance of marine
phytoplankton. However, collection of sufficient quantities of pure phytoplankton
from natural waters for radiochemical analysis is difficult. We therefore approachedthe
problem by experimentally studying the accumulation of U-Th series nuclides in different species of marine phytoplankton
maintained in laboratory cultures. We have
included Synechococcw sp., a picoplanktonic cyanophyte. The picoplankton, relatively little studied to date, have been shown
to be an'important component of phytoplanklon communities (Platt et al. 1983)and
may significantly influence the movement
ofparticle reactive metals in oceanic surface
waters (Fisher 1985).
We thank R. Anderson, M. Bacon, J.
Cochran, D. DeMaster, S. Heussner,and P.
131
r32
Fisher et al.
Santschi for comments on an early draft of
this paper. R. Rengarajan and N. Hussain
helped with computer programming.
Materials and methods
Experiments assessingthe accumulation
of lJ, Th, Ra, and Pb in algal cultures used
the centric marine diatom Thalassiosira
pseudonana (clone 3H), the chlorophyte
Dunaliella tertiolecta (clone Dun), the coccolithophore Emiliania huxleyi (clone MCH
No. l), the filamentous cyanophyte OscilIatoria woronichinii (clone Osc N4), and the
picoplanktonic cyanophyte Synechococcus
sp. (clone Ll6O2). Cell dimensions, experimental procedures for media preparation,
sampling, and isotope handling are all described elsewhere(Fisher etal. I983a; Fisher 1985). Cultures were incubated in constant light or constant darkness. Two'
replicate cultures (three replicates for clone
Ll602) and appropriate controls, with no
cells, were examined for each treatment.
Volume/volume concentration factors
(VCFs) were calculated(Fisheret al. 1983a)
for all nuclides and algal species at each
sample time. The activity per cell volume
was calculated by subtracting appropriate
control filter values (from uninoculated
"cultures") from the measuredfilterable activities in the inoculated samples.The Synechococcussp. culture was axenic, but the
other cultures had some bacterial contamination as indicated by epifluorescencemicroscopy. However, in none of the cultures
did the bacterial biomass exceeda few percent of the algal biomass. Moreover, most
of thesebacteria passthrough the 1-g.mNuclepore filters used to filter the nanoplankton.
Interference from multiple gamma emissions during measurementswas minimized
by conducting experiments with either a
single tracer [232Uin equilibrium with its
daughters (seeFig. /), or ztoPbl or a double
tracer (2toPb and 228Ra)per culture. All
tracers were added as nitrates. No attempts
were made to determine the speciation of
the tracers after their introduction to the
seawater. The time-zero water concentrations in different experiments ranged between:232U,1.2 and 11 pM (0.23-2.1Bq
and 0.8 pM (0.6-5.7Bq
ml-t); 228Th,0.085
0.43 pM (1 Bq ml-t); ttoPb,
ml-t); 228Ra,
15 pM (10 Bq ml-t). Surfaceseawaterconcentrations of U are about 10apM, of Th
<0.3 pM, of Ra about 2 x l0-o pM, and
of Pb about 100 pM (Cochran 1982; Bruland 1983). The 232U'and 2r0Pbwere obtained from the C.E.A., Gif-sur-Yvette,
was milked {iom 232Th
France,and the 228Ra
and was provided by R. Bojanowski. The
pH of the cultures immediately after tracer
addition was between7 .4 and 8.0.
An additional experiment was conducted
to measure the retention of Th and Pb in
zooplankton fecal pellets. Several hundred
brine shrimp (Artemia salina\ were fed dia"double-latoms (clone 3H) that had been
228Th
2toPb
for 3 d and then
and
beled" with
resuspendedinto unlabeled filtered seawater (Fisher etal.1983a). After 6 h of feeding
on a suspensionof 2 x 105cells ml-r, the
animals were transferred to a liter of unlabeled filtered seawatercontained in a modified fecal pellet collector (La Rosa 1976).
The average fecal pellet produced was
361195 pm long and 148+35 s.m in diameter (measurementsof 80 pellets).
The pellets were gently collected by largebore pipet and placed on a 43-pm nylon
mesh fixed to the bottom of a polyethylene
tube of 1-cm length and diameter. The polyethylene capsulecontaining the pellets (50100 pellets per capsule) was then sealed at
the top with 43-pm nylon mesh. Twenty
such capsules were made and assayed for
their 46 keV radiation from 2toPb within
about 2 h ofcollection. Three ofthe capsules
were preserved as controls in a counting
tube, and the remainingcapsules were held
immersed in filtered seawatercontained in
a glass beaker at 4"C+ 1o or 14oC+1ofor
various periods up to 120 d. The water in
the beaker was changed at intervals of 1-3
d. The nuclide retention was measured followingtwo approaches.In the first, a marked
capsule from the set was taken out of suspension at frequent intervals, assayedfor its
2toPbactivity, and resuspendedin fresh, filtered seawater.In the secondapproach, one
of the capsuleswas taken out of suspension
permanently at defined intervals. This capsule was immediately assayedfor its 2r0Pb
activity and after a few months for its 228Th
activitv.
133
Th, Pb, U, and Ra in algae
I
zovr
I
I
rstvr
Fig. l.
I
I
:.uo I
-P;l
-ffi1'-*J
I to.on|
|
| oomtn
|
staore
I
2!2Udecay series.Nuclides with half-lives > I h are shown.
The concentrations ofradionuclides in all
samples(i.e. water, filters,andArtemiafecal
pellets)were measuredby alpha and gamma
spectrometric methods. 232U'was measured
by alpha spectrometry with 238U:as tracer.
IJranium was separated from the samples
and purified by standard ion exchangeprocedures, electro-deposited, and measured
with a solid-statedetector(Bhat et al. 1969).
""Th, ttoPb, and 228Rawere measuredby
nondestructive gamma ray spectrometry.
228Thwas determined by counting its progeny, 2t2Pb(22U260 ke$, to avoidthe timeconsuming analytical separations for puri228Th. The intermediate daughters
fuing
between 228Thand 2r2Pbare all short-lived,
except fbr 224Rawhich has a half-life of 3.64
d (Fig. 1). The gowth of 2t2PbIiom 228Th
would be governed by the Bateman equation (Friedlander et al. 1966) and to a large
extent dictated by the half-life of 22aRa.The
2r2Pband 228Thactivities would nearly attain radioactive equilibrium after about 56 half-lives of 22aRa(around 20 d) and the
2t2Pbactivity of the sample after this period
would be nearly the same as that of 228Th.
(This approach would not be valid if the
uptake of Ra in the cultures exceededthat
of Th, although none of our samples falls
in this category.)
We counted the samplesperiodically over
15-20 d and calculated the 228Thand 22aRa
activities in the samplesat the time of their
filtration by analyzing ztzpb time-series
count rates with the Bateman equation. The
calculations assume that there is no loss of
22oRnfrom the samples. The observation
that in each of the 2O time-zero samplesthe
2t2Pb count rate remained nearly constant
(C.V. : 3-7o/o)throughout the 20-d counting period suggeststhat 22oRnloss from the
sampleswas negligible. 2toPbwas measured
by counting its 46 keV photon emissions
(35-55 keV) and 228Ravia the photons of
its daughter 228Ac(865-1,010 keV). The
228Rasampleswere countedl-Z d after their
filtration to ensure radioactive equilibrium
between228Raand 228Ac(t,^: 6.1 h). All
phytoplankron samples were counted with
a Packard 5650 Autogamma counter [with
a NaI(T1) crystall under identical counting
conditions. The 228Thassay (via 2t2Pb)in
the fecal pellets was performed with an Ortec intrinsic gamma-X germanium detector.
Whenever a double tracer was used, necessarycorrections for interference in the peak
regions were made by counting appropriate
standards.
Results
The growth of the cells and their accumulation of 228Thand 2r0Pbare shown in
Figs. 2 and 3. The uptake of 232U'and228Ra
by the phytoplankton was extremely low
(<2o/o of the activity in the water column,
throughout the course of the experiments).
The fractions of filterable radionuclide in
the uninoculated control sampleswere very
low (< lolo)for all nuclides except 228Th,for
which it averagedabout 230lowith the nanoplankton experiments and < 100/owith the
picoplankton experiment (Figs. 2, 3).
Equilibrium with respect to fractionation
between the dissolved and the particulate
phases of the radionuclides was generally
achieved in 1 d. The nuclide concentration
per cell at 1 d was generally within a factor
of two of its value after a 3-d exposure
for all isotopes and algal speciesexcept for
the 2roPbcontents per 3H cell in the light
after a 3-d exposure, where it was 5-6 times
lower than those after exposurefor I d. This
observation probably resulted from a cell
division rate of 3H that exceededthe 2roPb
adsorption rate. Thus, the VCF values did
not change appreciably after I d, except for
2toPbin illuminated 3H cells (Figs. 2, 3).
Generally, algal uptake followed the pattern
228Th;, zropf 2 z3zlJ - 228Ra.Concentration factors after exposurefor 3-6 d are given in Table 1. The differencesbetween the
VCF values.for the different algal species
Fisher et al.
t34
log cells ml-1
228Th uptake
log VCF
3H
A
Dun
|
----4
{
Mch 4
Osc
4
Syn 5
DAYS
DAYS
2
DAYS
Fig.2. Growth of cells (3H: Thalassiosirapseudonana;Dvn: Dunaliella tertiolecta;Mch: Emiliania huxleyi;
22tTh under light (a) and dark (O)
Osc;-Oscillatoria woronichinii; Syn: Synechococcussp.) and accumulation of
22tThcaughton a I -pm Nuclepore
22ETh
percent
water
column
of
the
uptake representsthe
conditions. The percent
filter (0.2-pm filier for Syn). 22tThuptake by control filters (from uninoculated water) shown with solid symbols;
22tThVCF values take into account the 22EThcontent of
controls for 3H, Dun, Mih, and Oic identical. The
blank (control) filters. Data points are means + I SD of two replicate cultures (three replicatesfor Syn).
r35
Th, Pb, U, and Ra in algae
log cells ml-1
21o
Pb uptake
log VCF
(%)
1
:
3H
5
5
\
4
6
4
6
Dun 5
4
6
Mch
DAYS
Fig. 3. As Fig. 2, but of 2toPb.
Fisher et al.
136
o)
.g
o
E
o
(I
(D
10
=
€(U
ro
4"C
14"c
0
1m
110
DAYS
22sTh
2roPb
(!,
(O,
O) and
Fig. 4. Retention of
) by Artemia salina feeal pellets maintained in unlabeled
seawaterat 4"C and l4qc.
for any one nuclide were smaller than the
differencesamong the nuclides for any one
species.The highest VCF value measured
was -2 x 106fbr 228Thand 2r0Pbin Synechococcussp. (Table 1). The VCF values
for 2roPband 228Thin algaeincubated under
light and dark conditions were generally
similar (Table 1; Figs. 2, 3).
In the illuminated D. tertiolecta and O.
woronichinii cultures, pH increasedconsiderably as the cells divided, going from about
7.9 after I d to about 9.0 after 3 d. Since
the VCF values for 2r0Pband 228Thin light
and dark cultures were similar (Table 1),
this pH increase apparently had no major
effect on nuclide uptake by these species.
The retention of 2r0Pbin A. salina fecal
pellets contained in the repeatedly sampled
capsulesis shown in Fig. 4. Both at 4oCand
14'C, -90o/o of the 2toPbin the pellets is
lost within 120 d. The loss curve at 4oCis
nearly exponential \Mith a single rate constant,which yields a retention half-time (rbu,)
of 36+3 d fbr 2toPbin the pellets.The loss
curve at I4"C seems to show two components, characteized by a rapid loss during
the initial stages(tb*: 10+0 d) followed
by a more gradual loss (lbu,: 32+4 d). If
we combine the 2r0Pbloss data from different capsulesfor the same temperature and
time, the loss curves show greater scatter;
the tbv,values for the pooled data are then
45+6 d at 4oC and 23+2 d at 14'C. The
fecal pellets retained 228Thmore than ttoPb,
as there was no discernible loss of 228That
4"C, while atl4"Cthetbr,was30-44 d. The
differential retention of 228Thand 2r0Pbin
fecal pellets held at 4"C suggeststhat physical breakup of the pellets was not solely
responsible for the measured loss of radionuclides (at least at 4"C).
Discussion
The concentration factors of the nuclides
in the algal species studied generally followed the trend Th > Pb > Ra = IJ. Comparison of the concentration factors with
those measured for transuranic elements in
these cells (Fisher et al. 1983a; Fisher and
Fowler in press) shows that the accumulation of Th is similar to the particle-reactive
elementsAm, Pu, Cm, and Cf. The minor
influence of light on the accumulation of
U-Th series nuclides in these species was
similar to earlier findings for other metals
by Fisher et al. (1983a, 1984) and Fisher
(1985).Theseearlier studiesconcludedthat
metal association with phytoplankton proceeds by adsorption to cell surfaces. Lead
137
Th, Pb, U, and Ra in algae
Table l. Volume/volum€ concentration factors (VCFs) obtained from phytoplanlton in culture. Means +
I SD, rz: 2 for all cellsexceptLI602 {n : 3)' (Not determined-nd.)
Clone
Condilion
22rThr
39+ll
106+23
3H
(diatom)
Light
Dark
Dun
(green)
Light
Dark
4l+7
5 7+ 7
Mch
(coccolithophore)
LiCht
Dark
3l+12
49+23
Osc
(blue-green)
Light
Dark
1.1+0.7
2 . 1 +l . l
Lr602
Light
Dark
l90r l2t
r99+24t
(blue-green)
'roPbt
3.7+0.4
4 7+ 6
l0r4
6.3+2
3.4+0.6
4.7+0.9
1.3+0.2
1.9+0.6
264+61
178+46
2!2(J*
<0.02
<0.06
<0.04
<0.02
<0.01
<0.06
<0.004
<0.002
nd
nd
,rRa+
<0.03
nd
nd
nd
<0.03
nd
<0.006
nd
nd
nd
* Equilibration time of 96 h.
f Equilibration time of 70 h.
* Equilibration timc of 144 h.
and americium have been shown to associate principally with the cell walls of marine phytoplankters (Fisher et al. 1983c),
and presumably other particle-reactive,
nonessential metals like Th would do the
same.
The high affinity of Pb and especially Th
for phyoplankton suggeststhat they are
particle-reactive in general, consistent with
their known marine geochemical behavior
and their long retention in Artemia fecal
pellets. The association of Th with marine
particulate matter has been shown to be reversible and in equilibrium with adsorption/desorption processes (Nozaki et al.
1981;Bacon and Anderson 1982).In open
ocean waters, the residencetimes of Th isotopes in marine suspendedmatter have been
calculatedto be in the rangeof 2-10 rnonths
(correspondingto a tb^of about 45-200 d),
based on vertical profiles of dissolved Th
isotopes in open oceanwaters (Nozaki et al.
l98l; Bacon and Anderson 1982).
We have observed that the tb* of 228Th
in Artemia fecal pelletsheld at 4C is > 120
d. The retention of 2roPband 228Thin the
fecal pellets was comparable to that of
24rAm-a particle-reactivetransuranic nuclide-in euphausiid fecal pellets (Fisher et
al. 1983b).The scatterin the 2roPbretention
data derived from different fecal pellets (i.e.
different capsules) probably was more a
function of different physical characteristics, including degradation rates, among the
different batches of pellets than of varia-
tions in the desorption rates for the nuclides. The influence of biological and physical degradation of the fecal pellets, if any,
on retention of radionuclides warrants further study. It can be surmised from our data
that the desorption of these elements from
sinking fecal pellets would proceed slowly
in nature, and fast-sinking fecal pelletswould
be expectedto transport theseelements,like
the particle-reactive transuranic elements,
to deep waters and sediments (Higgo et al.
1977;Fisher and Fowler in press).Consistent with this idea, Coale and Bruland (1985)
concluded that zooplankton fecal pellets
were responsible for the transport of Th in
the California current.
Table 2 presents a comparison of concentration factors of Th, Pb, Ra, and U in
the phytoplankton cultures with those in
various types of suspended particulates in
seawater. Concentration factors were also
calculated on dry weight bases(DWCFs) to
facilitate comparisons with available field
data. The geometric mean DWCFs in the
phytoplankton cultures studied here are
8.7 x 105for Th and 2.9 x 105for,Pb. If
we exclude the D. tertiolecta and O. woronichinii data to avoid any possible pH artifacts, thesegeometric means are not altered
appreciably (in fact, they increase about
twofold). The geometric mean DWCFs in
phytoplankton cultures are comparable to
those measured in surface ocean particulates and in sediment trap materials (Table
2). Field data of Krishnaswami et al. (1976)
Fisher et al.
138
Table 2. Comparison of dry weight concentration factors (DWCFs) obtained from laboratory culture experiments with those determined from measurementsof marine particulate matter.
DWCF (x l0')
Algal cultures
T. pseudonana
D. tertiolecta
E. huxleyi
O. woronichinii
Synechococcussp.
Surfacewater particles
Atlantic Ocean
Indian Ocean
GeosecsNo. 48, S. Atlantic
Deep water particles
Panama and Guatemala basins
Sediment trap material
N. Atlantic (389 m)
N. Atlantic (1,000m)
Net plankton
Phytoplankton
Phytoplankton
Mixed plankton
Mixed plankton
Zooplanhon (copepodsand
crab larvae)
Zooplankton (salps)
Zooplanhon (calanoidsand
cyclopoids)
106
244
89
2.2
975
l0
60
l0
3
1,100
<0.08
<0.09
<0.01
<0.06
<0.2
<0.03
<0.008
2
200
80
Krishnaswami et al. 1976
Krishnaqwamiet al. l98l
Somayajulu and Craig 1976
l0
2,000
200
450
Bacon and Anderson 1982
50
100
:
0.6
2
Present work
Present work
Present work
Present work
Present work
3
7
Brewer et al. 1980
Brewer et al. 1980
z;a
-
0.18
0.7
0.12
Shannonet al. 1970
Shannon and Cherry l97l
Szabo 1967
Knauss and Ku 1983
0.01
0.03
Krishnaswami et al. 1985
Krishnaswami et al. 1985
0.02
Kharkar et al. 1976
0.2
l
indicate that 23aThconcentration factors in
marine particulate matter decreasewith increasing suspendedparticulate load. Since
the particulate load in ourexperiments (>0.3
mg liter-t) exceededthat of all but the most
productive waters, it may be expected that
the measured concentration factors (and
subsequent flux estimates) in the phytoplankton culturesrepresentlower limits. The
high concentration factors of Th and Pb in
phytoplankton suggestthat these organisms
could be important sequesteringagents for
thesenuclides in surfacewaters and that the
flux of organic material out of euphotic
waters, either by direct sinking of phytoplankton cells or packagedinto zooplankton
fecal pellets, could appreciably influence the
vertical flux of these nuclides.
Table 3 presents a range of estimates of
organic carbon exported from the euphotic
zones of different oceans together with an
estimate,basedon mean DWCFs ofTh and
Pb derived in our laboratory experiments,
0.05
0.05
0.5
of the flux of 234Thand 2roPb associated
with sinking phytoplanl<ton (in whatever
form). The flux of radionuclides should be
a function of the sinking particulate load,
which is in turn a function of the primary
production in the euphotic zone (Honjo
1982;Deuseret al. 1983;Jickellset al. I 984),
provided that the radionuclide: C ratios in
the primary producers and settling organic
matter are identical. However, if the radionuclides are not recycled in surface waters
as efficiently as carbon, then the nuclide
fluxes presentedin Table 3 would be lower
Iimits. The estimates of 234Thand 2roPb
fluxes on an oceanwide basis via sinking
phytoplankton (Table 3) can be compared
with the measured fluxes of these nuclides
from the euphotic zone.
Data on the 23aThremoval rate from surface waters are obtained from 234Th' 238U'
ratios. On the basis of a mean value of 0.8
for the 234Th' 238U.activity ratio in surface
waters (- 100 m) of the mid-Arabian Sea,
Th, Pb, U, and Ra in algae
Wharton Sea, and North Pacific, the 234Th
removal is calculatedto be -50 x 10adpm
m-t yr-t (Bhat et al. 1969; Matsumoto
1975). For oligotrophic waters such as the
SargassoSea,an estimateof 234Thremoval
can be obtained from 228Thdata. The residencetimes of 228Thare (6 yr in the upper
100 m of these waters and <29 yr in the
upper 350 m, correspondingtoremoval rate
constantsof >0.167 yr-t and >0.035 yr-t
(Li et al. 1980).If we assumethat 234Thand
228Thhave the same residencetime (Kaufman et al. 1981) and a234Thstanding crop
of 2 x 105 dpm --z (at 100 m), then the
234Thremoval flux can be calculated to be
0.7 to 3.3 x 104dpm 6-zy1-t. We estimate
the removal flux of 234Thin oligotrophic
watersvia sinking phytoplankton to be l.l x
l0a dpm m-2 yr-t, using a new production
rate of I.54 gC m-2yr-r (Eppleyand Peterson 1979) and the mean DWCF of 8.7 x
105in phytoplankton cultures.
For waters of high productivity such as
in the Antarctic, there are no water-column
data for calculating the 234Thremoval flux,
although an upper limit can be set by assuming the removal rate = production from
139
Table 3. Estimates of fluxes of 234Thand 2'0Pbinfluencedby phytoplankton in different waters.
Flux from the euphotic zone
Org. C'
GCm-'
yr-')
Atlantic
27
Pacific
8
Indian
18
Arctic
0
Antarctic
146
Flux at - 100 m
:r.Tht
( x 1 0 .d p m
m-2 yr-r)
18.8
5.6
t2.5
0
r02
O.z-3.3$;
50ll
ir0Pb+
( x 1 0 3d p m
m-2yr-r)
6.3
1.9
4.2
U
33.9
5-20#
t Assumesequalto "new" primary production(from EppleyandPctcrson
1979).
t Using geomerric mean DWCF of 8.7 x I 0t (Table 2) and surfacewater
2r"I'hof 2 dpm liter-r(Coalcand Bruland 1985);assumes
C:0.25
dry wt.
x
gcometric
(Tablc
mean DWCF of 2.9 lOt
2) and surfacewater
* Using
2roPbconcentrationof 0.2 dpm liter-r (Cochran 1982);assumcsC wt.
0.25 dry
0 For oligotrophicwaterssuchas the SargassoSea(from Li ct al. I 980);
assumes2!'Th and 22rThrcmoval rale constantsarc equal (Kaufman
e rs l . l 9 8 l ) .
ll Calculatcdfrom Bhat et al. 1969 and Matsumoto 1975.
# From Bacon.l al. 1976and Nozaki and Tsunogai 1976.
parison of measured 234Thfluxes at specific
oceanic sites with fluxes predicted for these
waters using our laboratory-derived concentration factors and available data on new
production rates shows good agreement.For
238U'ry 200 x 104 dpm p-zy1-t.
For 2t0Pb, example,in the California Current (36o50'N,
x
123"00'W; MLML Pit Cruise-2) our data
the removal flux is about 5-20
103dpm
sl-z y1-t (Table 3). A major problem asso- would predict a removal flux of 8 x 105
ciated with calculating 2r0Pb removal flux dpm m-2 yr-r for a new production rate of
is the paucity of data on its atmospheric I l6 g C m-2yr-'(Coale and Bruland 1985).
deposition. When estimated234Thand 2roPb This estimate compares with the value of
removal fluxes via phytoplankton are com- -7.3 x 105dpm 6-z y1-t basedon water
pared with those calculated from their sur- column inventories and sediment trap flux
face water (-100 m) inventories, both in of 234Th(Coale and Bruland 1985). Simioligotrophic waters (suchas the central sub- larly in the VERTEX-I station (36'36'N,
123%8'W) our data would predict a 234Th
tropical regions) and in hiehly productive
(e.g.
waters
upwelling regions,neritic waters, removal flux of 5.7 x 105dpm m-2yr-r for
Antarctic waters), we find that > 500/oof the a new production rate of 82 g C rys-zy1-t
234Thsedimenting out at 100 m probably is (Knauer et al. 1984). The 234Thflux at this
associated with the phytoplankton or their location basedon its deficiency in the water
debris. Further, the estimates in Table 3 column is 5.5 x 105dpm m-2 yr-t (Coale
show that the phytoplankton and their de- and Bruland I 985). For 2r0Pba similar combris could be a significant pathway for re- parison can be made for the MLML Pit
moving 234Thand 2r0Pbfrom surface ocean Cruise-2station.Basedon a DWCFof2.9 x
waters.
105,a new production rate of 116 g C m-2
yr-r,
and a 2roPbconcentration of 0.2 dpm
Support for these suggestionscomes from
(Bacon
studies
et al. 1985; Coale and Bru- liter-r, we would predict a removal flux of
land 1985) showing positive correlations 2.7 x L04dpm m-2yr-t. The measuredflux
between 228Thand 234Thfluxes and primary at this station is 1.9 x lOa dpm 6-z y1-t
production in various surface waters of the (Coaleand Bruland 1985).
North Atlantic and the North Pacific. ComIn light of these findings, the suggestions
140
Fisher et al.
of Nozaki and Tsunogai (1976) and Tur- highly calcified macroalgae, presumably by
ekian (1977)that biologicalremoval of 2toPb coprecipitation with the calcium carbonate
matrix (Edgingtonet al. 1970).However, in
and 23aThwould account only for a minor
fraction of their flux from surface waters our study no pronounced differences were
need re-evaluation. The lower 2r0Pb and observed between the concentration of Ra
234Thenrichment factors these workers re- in the coccolithophore,E.huxleyi and in the
ported for plankton probably result from
noncalcareous phytoplankton. Analogous
measurementsof comparatively large, net to Ra, the distribution of Ba in the sea applankton, dominated by zooplankton. We pears to be linked to that of Si (Li et al.
would speculate that particle-reactive ele- 1973; Collier and Edmond 1984). As with
ments like Th and Pb should show a direct Ra, the major siliceouscarrier phasefor Ba
correlation between enrichment factors and hasnotbeen identified @ankston etal.1979;
surface: volurne ratio of the particles (or or- Collier and Edmond 1984). It is likely that
ganisms),as seenfor Am, Cm, Pu, and Cf the cycling of Ra and Ba are governed by
(Fisher and Fowler in press), thereby re- the same process(es),such as the biologisulting in higher concentration factors in the cally mediated formation of barite crystals
smaller phytoplankton than in the larger in the water column (Dehairs et al. 1980).
phytoplankton and the zooplankton. The
The low concentration of U in the phydata for Th and Pb (this paper) and other toplankton may be a result of the speciametals (Fisher 1985) in the picoplankton tion of U in an anionic complexfurther support this idea. The data of Shan- UOr(CO3)ro- -in seawater (Stumm and
non et al. (1970)on 2r0Pbin net phytoplank- Brauner 1975).This complex may be genton yield a DWCF of I x 105,comparable erally unreactive for suspendedmarine parto the mean value based on our experi- ticulates (such as phytoplankton) which carments. It is recognized,however, that there ry a negative surface charge (Myers et al.
is considerable scatter in the available field
1975; Neihof and Loeb 1972). Moreover,
data.
the affinity ofU for dissolved carbonate may
The close similarity in the dissolved pro- greatly exceedits afrnity for ligands on algal
files of 226Raand Si in the world ocean sug- surfaces. Consistent with this hypothesis,
gests that 226Radistribution in the oceans algal uptake of IJ, which was attributed to
is affected by the same processesthat affect U adsorption to cell surfaces,was found to
the Si distribution (Ku et al. 1970, 1980; be maximal at pH 5, being markedly inhibChung 1980). Sediment trap data on the ited by the presenceofcarbonate (Sakaguchi
settling fluxes of 226Raand biogenic SiO2 et al. 1978;Horikoshi et al. 1979).Analyses
show a very strong positive correlation with
of sediment trap samples yield a IJ concena slope of 26 dpm 226Rsg-t SiO2 (Brewer tration of about 1.1 dpm g-t dry biogenic
et al. 1980). Given a typical 226Raconcen- matter (Breweret al. 1980;Anderson 1982),
tration of 0.2 dpm liter-t in seawaterand a corresponding to a DWCF of about 500.
biogenic SiO2 content of 50o/oin siliceous This value is consistent with the results oborganisms, the sediment trap data yield a tained from the phytoplankton cultures (TaDWCF of about 6 x 104in siliceousorgan- ble 2), suggesting that the phytoplankton
isms. Attempts to identifu 225Ra-concen- uptake of U in surfacewaters could account
trating organisms have thus far not been for the measured fluxes of this element
successful,exceptfor the observation of high through the water column.
concentrations of 226Rain Rhizosolenia sp.
In summary it would appear that Th and
(Shannonand Cherry I97I). Our resultson Pb are very reactive for marine particulate
Ra uptake by T. pseudonana, another cen- matter, including phytoplankton and zootric diatorn, failed to show substantial plankton fecal material. Their flux from surbioaccumulation of this element. Ra has face waters should be governed by associbeen shown to be concentrated heavily by ation with the surfacesof sinking particulate
zooxanthellae associated with coral reefs, matter and therefore be a function of priwith a DWCF of about 3 x 106(Flor and mary productivity in the euphotic zone. By
Moore 1977).Similarly, Ra concentratesin contrast. U and Ra behave relativelv con-
Th, Pb, U, and Ra in algae
t4l
nuclides, p. 384-430. In M.
U- and Th-series
servatively with respect to association with
Ivanovich and R. S. Harmon [eds.], Uranium sebiogenic particulates. We note, however,
ries disequilibrium - applications to environmenthat these conclusions are based on expertal problems. Oxford lJniv. Press.
imental laboratory studies with only a few Collren, R., aryu J. Ept"toNp. 1984. The trace element geochemistryof marine biogenic particulate
algal species.It is difficult to judge how far
matter. Prog. Oceanogr.13: I l3-199.
these laboratory culture experiments can be DeHerns, F., R. CnrssELET,AND J. Jnpwnn. 1980.
used to predict natural marine biogeochemDiscrete suspendedparticles ofbarite and the barium cycle in the open ocean. Earth Planet' Sci.
ical processes.However, the observation
Ixtt. 492528-550.
that our results are consistent with available
W. G., P. G. Bnewen, T. D. Jlcxu.Is, nNp R.
DEUsen,
field data is certainly encouraging. Further
F. Couumu. 1983. Biological control of the reresearchwith many more speciesand varymoval ofabiogenic particles from the surfaceocean.
ing experimental conditions should help to
Science219: 388-391.
quantifu the role of organisms in mediating EocrNGroN, D. N., S. A. GonooN, M. M. Tnoulrars,
nNp L. R. At-uopovm. 1970. The concentration
the distribution of these nuclides in marine
of radium, thorium, and uranium by tropical masystems.
rine algae.Limnol. Oceanogr.15: 945-955.
References
ANoeRsoN,R. F. 1982. Concentration, vertical flux,
and remineralization of particulate uranium in
seawater.Geochim. Cosmochim. Aeta 46212931299.
BecoH, M. P., mrp R. F. ANornsoN. 1982. Distribution ofthorium isotopesbetweendissolvedand
particulate forms in the deepsea.J. Geophys.Res'
87:2045-2056.
-,
C.-A. HuH, A. P. Furn, lNo W. G. Deusrn.
1985. Seasonality in the flux of natural radionuclides and plutonium in the deep SaryassoSea.
Deep-SeaRes. 32: 273-286.
D. W. SpsNcER,AND P. G. BnewER' 1976.
2roPb/226Ra
disequilibria in seaand 2r0Po/2r0Pb
water and suspended particulate matter. Earth
Planet. Sci. I*tt. 322277-296.
AND
Ber.ucsroN,D. C., N. S. Flsnrn, R. R. GurLL,c,RD,
V. T. BowrN. 1979. Studiesof elementincorporation by marine phytoplankton with special
referenceto barium. Environ. Meas.I-ab. Environ.
Q.U.S. Dep. EnergyEML-349,p. 509-531.
Bn,tr, S. G., S. KrrsxNAswAMl, D' Ler,, RAMA, ANo
ratios in the
W. S. Moons. 1969. 234TW238U
ocean.Earth Planet. Sci. Irtt. 5: 483-491.
4. t'
BnswEn,P. G., Y. NozAKl, D. W. SpeNcER,,AIID
Fleen. 1980. Sediment trap experimentsin the
deep North Atlantic: Isotopic and elementalfluxes.J. Mar. Res. 38: 703-728.
BnoecKEn, W. S., A. K.lurrtalN, ANo R. M. Tmsn.
1973. The residencetime of Th in surfacewater
and its implications regardingfate of reactive polIutants. Earth Planet. Sci. Lett. 20235-44.
nNp T.-H. Pmc. 1982. Tracersin the sea.
Eldigio.
BnuLANo, K. W. 1983. Trace elementsin sea-water,
p. 157-220. In J. P. Riley and R. Chester [eds.],
Chemical oceanoglaphy,v. 8. Academic.
CHUNc,Y. 1980. Radium-barium-silicacorrelations
and atwo dimensional radium model forthe world
ocean.Earth Planet. Sci. Lett.49: 309-318.
Coar-e,K. H., ruto K. W. Bnu-aNp. 1985. 23oTh:
23sUdisequilibria within the California current.
Limnol. Oceanogr. 30222-33.
CocHneN. J. K. 1982. The oceanic chemistry of the
Errrev, R. W., eNo B. J. PsrsRsoN. 1979. Particulate
organic matter flux and planktonic new production in the deep ocean. Nature 2822677480.
FrsHrn, N. S. 1985. Accumulation of metals by marine picoplankton. Mar. Biol. 87: 137-142'
P. BrennecAARD,axp S. W. Fowun. 1983a.
Interactions of marine plankton with transuranic
elements.l. Biokinetics of neptunium, plutonium,
americium and californium in phytoplanlton'
Limnol. Oceanogr.28t 432-447.
marine plankton with transuranic elements. 3.
Biokineticsofamericium in euphausiids'Mar. Biol.
752261-268.
-r
M. BoxE, erp J.-L. Teyssri. 1984. Accumulation and toxicity of Cd, Zn, Ag, and Hg in
four marine phytoplankters. Mar. Ecol. Prog' Ser.
1 8 :2 0 1 - 2 1 3 .
K. A. BunNs,R.D. CHennv,ANoM.HsvRAup.
1983c. Accumulation and cellular distribution of
"tAm, 2r0Poand 2toPbin two marine algae'Mar.
Ecol. Prog. Ser. 11: 233-237.
-r
eNp S. W. Fowr"en. In press. The role of
biogenic debris in the vertical transport oftransuranic wastesin the sea. In T . P. O'Connor et al.
[eds.], Oceanic processesin marine pollution, v.
2. Krieger.
Fron, T. H., ANDW. S. Moonn. 1977. Radium/calcium and uranium/calcium determinations for
westernAtlantic coral reefs,p. 555-561. InPtoc.
3rd Int. Coral Reef Symp. Miami, Fla.
eNo J. M. MrLLen.
FnrrDlnNpsn, G., J. W. K-EI.u.{Epv,
1966. Nuclear and radiochemistry, 2nd ed' Wiley.
Hrcco, J. J., R. D. CHrnnv, M. Hrvuuo, aNp S' W.
Fowrrn. 1977. Rapid removalofplutonium from
the oceanicsurfacelayer by zooplankton fecalpellets. Nature 266t 623-624.
HoNro, S. 1982. Seasonalityand interactionof biogenicand lithogenic particulate flux at the Panama
Basin.Science218: 883-884.
HomosHr, T., A. NaxllIue, exp T. Slx.ecucm.
1979. Uptake of uranium by Chlorella regularis.
Agric. Biol. Chem. 43:617423.
Jrcrrus, T. D., W. G. DeusEn,4ryp,A.H. KNp. 1984.
The sedimentation rates of trace elementsin the
t42
Fisher et al.
234Th-238{J
radioactive disSargassoSea measured by sediment trap. Deep- Mnrsurraoro, E. 1975.
equilibrium in the surfacelayer of the ocean.GeoS e aR e s .3 l : 1 1 6 9 - 11 7 8 .
chim. Cosmochim.Acta 39:.205-212.
A., Y.-H. LI, e.NoK. K. TunrrrnN. 1981.
KAUFMAN,
The removal rates of 23aThand 226Thfrom waters Mvrns, V. B., R. L. IwnsoN, eNo R. C. Hnnmss. 1975.
The effectof salinity and dissolved organic matter
of the New York Bight. Earth Planet. Sci. Lett. 54:
on the surfacecharge characteristicsof some eu385-392.
KHenx-m, D. P., J. THousoN, K. K. TunEruAN,AND
ryhaline phytoplankton. J. Exp. Mar. Biol. Ecol.
17: 59-68.
W. O. Fonsren. 1976. rJranium and thorium
decayseriesnuclides in plankton from the Carib- NrrHor, R. A., nxp G. I. Lors. 1972. The surface
chargeof particulate matter in seawater.Limnol.
bean. Limnol. Oceanogr.2lz 294-299.
Oceanogr.l7z 7-16.
Krnusn, G. A., J. H. M.lnrrn, eNp D. M. I(nnr. I 984.
The flux of particulate organic matter out of the Noznn, Y., Y. HoRran,eNp H. Tsusora. 198l. Water
column distributions of thorium isotopes in the
euphoticzone, p. 136-150. In Global oceanflux
westernNorth Pacific. Earth Planet. Sci. Lett. 54:
study. Workshop Proc. Natl. Acad. Sci.
203-216.
Kxnuss, K., avp T.-L. Ku. 1983. The elemental
er.ro S. TsuNooru. 1976, 226Ra,2r0Pband
composition and decay-seriesradionuclide con- -s
2'0Podisequilibria in the western North Pacific.
tent of plankton from the East Pacific. Chem. Geol.
Earth. Planet. Sci. Irtt. 32t 313-321.
392125-145.
KnrsrrNeswevr,S., M. BasxnnaN, S. W. Fouen, nNo Pr,err, T., D. V. Sunsn RAo, eNp B. InwrN. 1983.
Photosynthesisof picoplankton in the oligotrophic
M. Hevnlup. 1985. Comparativerole of salps
ocean.Nature 30lz 702-704.
and other zooplankton in the cycling and transport
Rc,l'rl,M. Koroe, eunE. D. Gor-psenc. 1961. Leadof selectedelementsand natural radionuclides in
210 in natural waters.Science134: 98-99.
the Mediterranean waters. Biogeochemistry l: 353SnmcucHr, T., T. HonrKosHI, aNp A. Nlxerurle.
360.
1978. Uptake of uranium from seawater by miD. LAL, nNp B. L. Souevalulu. 1976. Incroalgae.J. FermentationTechnol. 56: 561-565.
vestigationsof gram quantities of Atlantic and PaSneNNoN,L. V., nr.ro R.D. Cnennv. 1971. Ra-226
cific surfaceparticulates. Earth Planet. Sci. ktt.
in marine phytoplanl:ton. Earth Planet. Sci. ktt.
32t 403419.
1r:339-343.
1981.
B.
L.
Souevnrulu.
M. M. Se,nrr*,
eNp
Chemicaland radiochemical investigationsof sur210 and lead-210 in the marine environment.
faceand deepparticles ofthe Indian Ocean.Earth
Geochim. Cosmochim.Acta 34: 701-711.
Planer.sci. I-ett. 54: 8l-96.
Ku, T. L., C. A. Hua, ANo P. S. CHeN. 1980. Me- Sounvlruru, B. L., ero H. Cnarc. 1976. Particulate
and soluble 2'0Pbactivities in the deep sea.Earth
ridional distribution of 226Rain the easternPacific
Planet. Sci. ktt. 322268-276.
Planet.
Sci.
along GEOSECScruise tracks. Earth
Sruuu, W., ANp P. A. Bn{rrrven. 1975. Chemical
Iett. 492293-308.
speciation,p. 113-239.In J.P. Riley and G. SkirY.-H. Lr, G. G. MenuEU, ANDH. K. Wowc.
row [eds.],Chemical oceanography,v. l. Academ1970. Radium in the Indian-Antarctic Ocean
ic.
south of Australia. J. Geophys. Res. 75: 5286Sznso, B. J. 1967. Radium content in plankton and
5292.
seawater in the Bahamas.Geochim. Cosmochim.
Le Rosn, J. 1976. A simple system for recovering
A c t a3 1 : l 3 2 l - 1 3 3 1 .
zooplanktonic fecal pellets in quantity. Deep-Sea
TuneKrAN, IC K. 1977. The fate of metals in the
Res.23:995-997.
oceans.Geochim. Cosmochim. Acta 4l: ll39eNo J. R. Toccwnnrn. 1980.
Lr, Y.-H., H. W. FEEt-Ey,
228Raand 228Thconcentrations in GEOSECSAttt44.
lantic surfacewaters. Deep-SeaRes. 27: 545-555.
-,
T. L. Ku, G. G. MerHrEu, A\rDIC WOT.GEMUTH.
1973. Barium in the Antarctic Oceanand impliSubmitted: 13 May 1985
cations regarding the marine geochemistryof Ba
226Ra.
Accepted:20 June 1986
19:
352-358.
Earth Planet. Sci Lett.
and