Limnol. Oceanogr.. 32(l), 1987, 189-200
0 1987, by the American Society of Limnology
and Oceanography,
Inc.
Oceanic stratified euphotic zone as elucidated by
234Th: 238Udisequilibrial
Kenneth H. Coale and Kenneth W. Bruland
Institute
of Marine
Sciences, University
of California,
Santa Cruz 95064
Abstract
Profiles of dissolved and particulate 234Th were determined at the VERTEX 2 and 3 stations off
Manzanillo, Mexico, and at the VERTEX 4 station about 900 km north of Hawaii. By modeling
the disequilibria between 234Th and 238U in the dissolved and particulate form, estimates of scavenging rates for Th from the dissolved to particulate phases, particle residence times, and the flux
of Th via particle removal can be obtained. 234Th: 238U activity ratio profiles indicate that the
euphotic zone can be separated into two layers: An upper oligotrophic layer characterized by low
new production values, low net scavenging, and long dissolved 234Th residence times; and a subsurface eutrophic layer with higher new production values, more intense scavenging, and shorter
dissolved 234Th residence times.
New production, rather than total primary production may determine net scavenging rates of
reactive elements from oceanic surface waters. These results contribute to the emerging descriptions
of the layered structure of oligotrophic euphotic zones and support the notion that this may be a
general and ubiquitous feature of global stratified oligotrophic regimes. These layered systems can
be structured not only in biological and nutrient parameters, but also in terms of the rates of
chemical scavenging and elemental transport.
The layered structure of biological processes within stratified oceanic euphotic
zones has been recognized for quite some
time. Since Steele and Yentsch’s (1960) paper on the vertical distribution
of chlorophyll, the subsurface chlorophyll maximum
has received a great deal of attention and is
now a well known oceanographic feature.
Dugdale (1967) proposed a two-layer model, for oceanic euphotic zones, in which nutrient-limited
phytoplankton
populations
overlie light-limited
populations. Venrick
(1982) reported two distinct floral assemblages associated with each of these layers,
which also seem to be differentiated ecologically in terms of competition and predation. Herbland et al. (1985) showed a
strong stratification in size classes of photosynthetic organisms (especially < 1 pm) of
the equatorial Atlantic, with the major shift
in size spectra occurring at or near the base
of the mixed layer. Recently, the presence
of stratified systems in meso- and oligotrophic regimes has been further interpreted in
terms of the concepts of Dugdale and Goering (1967) and Eppley and Peterson (1979)
regarding “new” and “regenerated”
pro-
duction (Harrison et al. 1983; Knauer et al.
1984; Jenkins and Goldman 1985). Such a
stratified euphotic zone can be visualized as
an oligotrophic layer overlying a eutrophic
layer which decays with depth to a classical
middepth aphotic system. After winter mixing, the euphotic zone becomes stratified in
terms of temperature, salinity, and oxidized
nutrients. The upper wind-mixed layer is
well lit and tightly coupled with respect to
nutrient cycling. By this we mean that practically all production within this layer is
thought to be maintained by recycling. In
this layer, new production is very low (about
5% of total production: Eppley and Peterson
1979), and nutrients supplied through winter mixing are depleted through the sinking
of large particles such as marine snow and
fecal pellets. In contrast, nutrient cycling in
that portion of the euphotic zone below the
mixed layer is less tightly coupled. There is
a greater supply of oxidized nutrients by
vertical mixing, and new production is much
higher in this region (about 30% of total
production). Consequently the flux of particulate organic carbon (POC) from this layer is much greater.
This stratified scenario, as proposed by
Knauer
et al. (1984) and Small and Knauer
’ This research was supported by National Science
(pers. comm.), is based primarily on meaFoundation grants OCE 79-23322, OCE 79-19928, and
OCE 82- 16672.
surements of primary production,
zoo189
190
Cbaie and Bruland
plankton grazing, fecal pellet production,
and sediment trap data. Evidence emerging
from these studies suggests that, although
the majority of the primary production generally occurs in the surface mixed layer of
oligotrophic waters, the POC flux is negligible immediately below this layer. In contrast, the euphotic zone below the mixed
layer has lower primary production; however, it appears to have substantial new production as evidenced by the POC flux at the
base of this zone. Thus the euphotic zone
during this time of year seems to be structured inversely with respect to new and total
primary production.
Previous studies in which :Z34Th: 238U disequilibria have been used to calculate residence times for dissolved and particulate
:134Th have shown dramatic: differences in
scavenging intensity and particle transport
processes between oligotrophic and eutrophic regimes. A natural extension of this
disequilibrium
approach is to the study of
scavenging intensity and particle transport
processes within a stratified system.
Numerous investigators have taken advantage of 234Th : 238U disequilibria to study
scavenging processes. Pioneering work with
total 234Th: 238U ratios was clone by Bhat et
al. (1969) and others (Matsumoto
1975;
Knauss et al. 1978; Kaufman et al. 198 1).
Subsequent investigators have determined
Th: U ratios in both dissolved and particulate forms (McKee et al. 1984; Santschi et
al. 1979; Bacon and Anderson 1982). More
recently Coale and Bruland ( 198 5) have also
demonstrated the utility of ;!34Th : 238U disequilibria for the eutrophic ICalifornia Current system, an area of minimal abiogenic
particle input. Using dissolved and particulate fractions, they showed that the scavenging of Th from dissolved to particulate
form varies as a function of primary production. They also suggeste:d that the rate
of particle removal was determined by the
types and abundances of grazing zooplankton present. In a similar manner Bruland
and Coale (in press) used 234Th : 238U disequilibria to determine variations in scavenging rate (dissolved to particulate)
in
surface waters throughout
a variety of
oceanographic regimes. These results indicate more rapid scavenging of dissolved Th
in areas of higher mixed-layer productivity
(coastal and equatorial divergence stations)
and much less intense scavenging in oligotrophic and mesotrophic regimes.
Whereas Coale and Bruland (1985) observed that the scavenging intensity of 234Th
was proportional
to total primary production within the California Current, Bruland
and Coale (in press) emphasized the dependence of scavenging intensity on particulate
organic carbon flux or new production
(rather than total production).
However,
vertical profiles in the oligotrophic regimes,
composed of relatively few samples from
the euphotic zone, failed to resolve any significant structure in scavenging rate within
the upper 100 m. Within the California Current system the euphotic zone is contained
within the mixed layer and 234Th : 238U disequilibria, as driven by primary production
or particle flux, are relatively uniform. These
234Th : 238U ratios increase toward equilibrium at depths just below the mixed layer.
New production is thought to vary as a function of total production and in similar ratios
throughout this regime (new production/total production = 20-30%: Eppley and Peterson 1979; Knauer and Martin 198 1). Because total primary production
and new
production covary in the California Current, 234Th scavenging could not be unequivocally related to either one of these
variables in this region. In contrast, in mesotrophic or oligotrophic
systems the euphotic zone may extend below the mixed
layer, dividing the euphotic zone into a well
lit upper layer where nutrients and carbon
are efficiently recycled and a lower layer
more characteristic of a coastal eutrophic
regime. A system that is structured with respect to total and new production and particle flux gives us a unique opportunity
to
evaluate the dependence of scavenging rate
on these parameters.
For simplicity we have chosen to treat our
stations as two-layered systems. If the scav-,
enging rate is ‘proportional to new production, which we will treat as being quantitatively equivalent to particle flux (over suf-,
ficiently long time intervals), then the region
of most intense scavenging, as determined
will OCCUI
by 234Th. * 238U disequilibria,
within the lower layer. On the other hand,
191
StratiJied Th : U disequilibria
is proportional
to total
than new production),
ost intense scavenging
50” N
ions on VERses off Manza4 in the central
10’ N
elation to new
I
Fig. 1. Station locations for VERTEX 2, October3, October-December
November
198 1; VERTEX
1983.
1982; VERTEX 4, July-August
one in more detail.
TEX 2 gave us our first indication of increased scavenging immediately below the
mixed layer. On VERTEX 3 and 4, we obtained more detailed profiles of the upper
water column.
We thank the crew and officers of the RV
Wecoma and Cayuse, G. Smith for assistance in VERTEX 4 sample collection, D.
Beals for assistance in radiochemical
processing and analysis of VERTEX 4 samples,
and G. Gill for comments on this manuscript.
Sampling and analysis
Seawater samples for this study were collected on three occasions from the VERTEX 2 and VERTEX 3 stations about 480
km west-southwest and 720 km southwest
of Manzanillo, Mexico, in the eastern tropical North Pacific and from the VERTEX
4 station about 900 km north of Hawaii
(Fig. 1). Shipboard sample processing was
used for 234Th to minimize in-growth and
decay corrections. Collections were made
with 30-liter, Teflon-coated
Go-Flo samplers (General Oceanics, Inc.). The Go-Flo
bottles were pressurized with filtered nitrogen gas, and the seawater was passed via
Teflon tubing through a 145-mm (diam),
0.3~pm pore-size Nuclepore membrane filters supported in a Teflon filter sandwich.
Dissolved samples were collected in 20-liter
Cubitainers and acidified with 50 ml of concentrated HCl. Filters for particulate analysis were stored frozen until processed.
For analysis, dissolved samples are spiked
with a 230Th yield tracer, equilibrated, and
coprecipitated with Fe(OH),. The Fe(OH),
fraction is collected by filtration on glassfiber filters and dissolved in HCl. Thorium
isotopes are separated and purified with a
series of anion exchange columns. Particulate samples are spiked with tracer and
processed with bomb digestion (Eggimann
and Betzer 1976) and subsequent anion exchange procedures similar to those used for
the dissolved samples. Final Th fractions
are electroplated on platinum disks. Thorium activities are determined by alphacounting with calibrated, silicon surface
barrier detectors used to determine yield,
followed by beta-counting with calibrated,
low background, gas flow detectors operated
in .anticoincidence with a common guard.
Uranium activities, collectively a conservative property of seawater (Turekian and
Chan 197 l), are not determined directly but
are calculated for each sample with the empirical relationship
238U dpm liter- ’ = 0.07081 x salinity
as determined by Ku et al. (1977). Using
these techniques, our analysis of four samples of deep water from the central North
Pacific, stored acidified for 1 yr, yielded
a mean 234Th: 238U activity
ratio of
1.008+0.013.
This is in excellent agreement with the expected equilibrium
value.
Scavenging model
Due to the relatively short half-life of
234Th, its high particle reactivity, and its
ubiquitous production by the alpha decay
of its conservative parent 238U, 234Th : 238U
Coale and Bruland
192
disequilibrium
has been shown to be an ideal tracer to quantify rates of particle scavenging and transport processes. The 234Th
half-life of 24.1 d and the analytical error
associated with current beta-counting techniques limit the utility of this tracer to processes with time scales between 1 and 300 d.
The 234Th: 238Uscavenging model of Coale
and Bruland ( 19 8 5) can be used to calculate
dissolved and particulate residence times
and the particulate flux for Th in oceanic
euphotic zones. Briefly, we exploit the fact
that the particle reactive daughter, 234Th, is
produced in nearly uniform
quantities
throughout the water column by the alpha
decay of its nonparticle reactive, conservative parent, 238U. We assume advection
and diffusion of 234Th to be negligible with
respect to vertical transport of 234Th via
sinking particles and that the activity of
234Th is not changing significantly over the
course of our observations (steady state). A
mass balance for Th in the dissolved form
can then be written:
JTh = A JTh - A“ThXTh.
(1)
and
rp = APTh/PT,.
(49
By assuming that the scavenging from the
dissolved to particulate form follows firstorder kinetics, we can define a scavenging
rate constant \kd, where
J Th
(5)
removal, we define
removal rate con-
= AdT,*d.
Similarly, for particulate
a suspended particulate
stant, !Pp,
PTh = APThap.
(61
Results
The locations of VERTEX 2, 3, and 4
stations are shown in Fig. 1; vertical profiles
for temperature, nitrate, plant pigments, and
dissolved, particulate, and total 234Th : 238U
activity ratios for the upper 300 m of the
water columns are presented in Fig. 2. The
234Th and 238U profiles for the upper 300 m
of the water column are presented in Table
1 together with the model-derived
param-
eters AdTh:AU, APTh:AU, AZTh:AU, T& \k&
TP, ep, JTh, and PTh. Occasionally, 234Th acAU is the activity of 238U, &., is the decay
tivities exceed 238U activities. In some cases
is therefore
this produced rather unusual (negative) valconstant for 234Th, and A&,
the rate of production of 234Th by the decay
ues for Td:,*& r,, 9,, JTh, and PTh. With respect to our model, negative values would
of 7-J. AdTh is the activity of dissolved
234Th, AdThXThis then the rate at which disimply a supply of 234Th other than from
solved 234Th is lost due to radioactive decay.
the decay of 238U. Such values may indicate
of particles
JTh represents the scavenging, or flux, of sinking and remineralization
carrying 234Th in a region close to secular
dissolved 234Th to the particles. For particthereby increasing total 334Th
ulate 234Th a similar mass balance can be equilibrium,
activities above the activity of 238U. It is
written:
our opinion that in most cases these negative
P,, = JTh - AD,&,.
(2) values are artifacts reflecting the analytical
uncertainty in attempting to discriminate a
JTh (from Eq. 1) is the source of particulate
system that is very close to equilibrium. The
234Th*ApThis the activity of particulate 234Th,
errors, given in Table 1 and depicted in Fig.
ApThiT,,is the rate at which particulate 234Th 2 represent +: 1 SD from the mean where
is lost due to radioactive decay and PTh is
replicate values (from sarnples or analyses)
the rate at which particulate 234Th is lost via
were available. For single determinations,
sinking or grazed particulates. Since JThand
analytical
uncertainty
was obtained by
PTh represent the non-radioactive removal
propagation of alpha- and beta-counting
rates for both dissolved (A”,,,) and particerrors and the corresponding uncertainties
ulate (APTh)234Th reservoirs, the mean res- in background, detector blanks, and effiidence times for 234Th can be defined for
ciencies as well as errors in all pipet delivboth the dissolved (Td) and particulate (TP) eries, yield tracer calibrations, volumes, and
phases:
time intervals used in calculating final 234Th
rd = Ad,dJn,
(3) activities.
193
StratiJied Th : U disequilibria
Temperature
(“Cl)
Nitrate
(PM)
Pigments (mg m -3)
0.1
Vertex
0.3
0.5
2
0.1
0.3
0.5
02
0.7
part.
0.2
-f,
1.0
0.6
diss.
0.6
tote
1.0
--an,-
4
Fig. 2. Vertical profiles of temperature, nitrate, pigments, and 234Th: *W ratios for VERTEX 2, 3, and 4.
For each station, upper horizontal lines delineate the base of the mixed layer and lower horizontal lines delineate
the base of the euphotic zone. The vertical line at 234Th : 238U = 1 represents radioactive equilibrium. The hatched
area represents total 234Th deficiency with respect to its parent 238U.
0.66kO.05
0.63kO.01
0.63 kO.02
0.65kO.03
0.49kO.01
0.42 kO.02
0.42-tO.02
0.56kO.03
0.89 kO.03
0.92 kO.04
0.85kO.03
0.83kO.04
0.82kO.03
0.84kO.04
0.8 1 kO.02
0.87kO.01
0.85 kO.03
0.88+0.0!
0.27kO.01
0.26kO.01
0.20+-0.01
0.24+0.01
0.58kO.01
0.60+0.02
0.4620.02
0.34+0.01
0.13-to.01
0.20+0.01
0.36zkO.01
0.35+0.01
0.36kO.01
0.23+0.01
0.12-+0.01
0.081kO.01
0.06kO.01
0.19+0.01
1.59kO.11
1.5 1 kO.02
1.53kO.06
1.58-f-0.07
1.2OkO.04
1.03kO.05
1.02kO.04
1.38-tO.06
2.18kO.08
2.2520.10
2.08 kO.08
2.04t-0.09
2.01 kO.08
2.06t-0.09
1.99kO.04
2.12kO.02
2.09kO.07
2.16kO.01
10
20
30
40
50
60
70
80
100
120
140
160
200
300
500
1,000
1,500
2,000
2.41
2.41
2.41
2.43
2.43
2.43
2.43
2.45
2.45
2.46
2.46
2.46
2.46
2.46
2.45
2.44
2.45
2.45
0.59kO.04
0.70+0.04
0.18+0.03
0.4 1 kO.03
0.83kO.03
0.93kO.07
0.94kO.06
0.82kO.06
0.80+-0.05
0.93kO.07
0.95a0.06
0.94kO.07
0.94t-0.06
0.99kO.08
0.59kO.02
0.98kO.02
0.94kO.04
0.3 l-10.02
0.3 1 kO.02
0.52kO.03
0.54kO.04
0.371kO.02
0.19+0.01
0.48kO.04
0.64kO.04
0.37kO.03
0.17+0.01
0.11 kO.01
0.07 rto.04
0.12+0.01
0.10+0.01
O.lOaO.O1
0.10~0.01
0.09~0.01
1.42a0.09
1.70+0.08
0.44kO.07
1.00+0.07
2.04kO.08
2.29kO.17
2.3OkO.16
2.03 +O. 14
l-97+0.12
2.26t0.16
2.31kO.15
2.3OkO.17
2.31k0.14
2.43kO. 19
1.45kO.08
2.4OkO.07
2.3 1 kO.09
5 2.41
30 2.43
45 2.43
60 2.43
75 2.44
100 2.45
125 2.46
150 2.47
250 2.46
500 2.45
750 2.44
1,000 2.45
1,500 2.45
2,000 2.45
2,500 2.45
3,000 2.45
3,500 2.45
50
82
7.7
25
175
482
510
161
139
430
599
568
581
3,375
50
1,436
575
68
59
60
65
34
26
25
45
280
382
191
168
156
182
151
230
200
255
VERTEX 2
0.72kO.04
0.83kO.04
0.39kO.03
0.64kO.03
0.99kO.04
l.OlkO.07
1.13kO.07
1.08-tO.06
0.95kO.05
1.00+0.07
0.99kO.06
0.97kO.07
0.99a0.06
1.03kO.08
0.63kO.03
1.02kO.03
0.98kO.04
VERTEX 3
0.77kO.05
0.74a0.01
0.71 kO.02
0.75kO.03
0.73kO.02
0.67kO.02
0.61kO.02
0.70+-0.03
0.94Iko.03
l.OOHLO4
0.99kO.03
0.97kO.04
0.97kO.03
0.93+0.04
0.86kO.02
0.90+-0.01
0.88-tO.03
0.96kO.01
0.128+0.007
0.1271kO.009
0.212+0.011
0.224+0.015
0.15 1 kO.008
0.075+0.006
0.196+0.014
0.260+0.016
0.151~0.011
0.07 1 kO.005
0.047+0.004
0.029+0.018
0.048+0.004
0.04OkO.003
0.039kO.006
0.04OkO.003
0.036+0.004
0.112kO.005
0.109+0.003
0.082+0.005
0.098-tO.004
0.239-+0.006
0.245kO.008
0.189+0.006
0.141 kO.006
0.053+0.003
0.082+0.004
0.144+0.005
0.142+0.005
0.147kO.006
0.094+0.004
0.051_+0.004
0.032~0.003
0.023kO.003
0.079+0.003
0.0147
0.0170
0.0167
0.0154
0.0296
0.0392
0.0397
0.0224
0.0036
0.0026
0.0052
0.0060
0.0064
0.0055
0.0066
0.0044
0.0050
0.0039
0.0202
0.0122
0.1307
0.0409
0.0057
0.002 1
U.OO2U
0.0062
0.0072
0.0023
0.0017
0.0018
0.0017
0.0003
0.0199
0.0007
0.0017
17
14
9.9
14
31
26
17
17
33
i ,747
537
166
145
50
13
11
6.3
68
16
26
12
21
346
-322
-51
-109
108
602
198
34
205
-47
3.7
-85
61
0.0577
0.0696
0.1009
0.0729
0.0322
0.0390
0.0594
0.0607
0.0307
0.0006
0.0019
0.0060
0.0069
0.0202
0.0775
0.0896
0.1577
0.0148
-0.0092
0.0093
0.0017
0.0050
0.0290
0.0049
-0.0213
0.2714
-0.0117
0.0165
-0.oi94
0.0641
0.0386
0.0824
0.0468
0.0029
-0.003 1
23
26
26
24
35
40
41
31
7.8
5.9
11
12
13
11
13
9.2
10
8.5
29
21
57
41
12
4.7
4.5
13
14
5.3
3.9
4.1
4.0
0.7
29
1.7
4.0
16
18
20
17
19
23
27
21
4
0.1
0.7
2.1
2.5
4.7
9.6
7.0
8.8
2.9
20
12
43
25
1.1
-0.6
-9.3
-5.9
3.5
0.3
0.6
2
0.6
-2.1
26
-1.2
1.5
Table 1. This table presents the measured, calculated, and model-derived
parameters. The units are as follows: Depth (m); A,, AdTh, Ap,,., (dpm
liter-l); the activity ratios are unitless; AzTn represents total (dissolved + particulate) 234Th; 74, 7P (d); *d, @,, (d-l); JTh,Pn (dpm 1,000 liter-l d-l). For
this model 9, and Qk, represent pseudo first-order scavenging rate constants and are described in the text, units are d-l.
?I
8
2
Au
2.50
2.49
2.49
2.48
2.48
2.47
2.47
2.46
2.46
2.45
2.45
2.43
2.41
2.44
2.45
Depth
(m)
25
40
50
60
75
90
100
125
150
175
200
250
500
1,000
1,500
2.51kO.04
2.31f0.11
2.04kO.07
1.36kO.05
1.6OkO.04
1.74kO.02
1.68kO.03
2.28-tO.00
2.17+0.10
2.51kO.10
2.3OkO.10
2.28kO.19
2.36a0.07
2.49kO.08
2.04t-0.07
-6-h
Table 1. Continued.
0.18kO.04
0.17f0.01
0.2Ot-0.01
0.38kO.01
0.47kO.01
0.33kO.01
0.33kO.02
0.20+0.01
0.26+0.01
0.17+0.01
0.13+0.01
O.OS+O.OO
0.09-t-0.01
0.11~0.01
O.ll+O.Ol
4-h
l.OO+-0.02
0.93kO.04
0.82kO.03
0.55kO.02
0.64kO.02
0.71f0.01
0.68kO.01
0.93+-0.00
0.88kO.04
1.03kO.04
0.94kO.04
0.94kO.08
0.98kO.03
1.02+0.04
0.83kO.03
A,
A,
Au
“Cl
VERTEX4
0.072+0.017
1.07kO.02
0.067+0.004
0.99kO.04
0.082kO.004
0.9OkO.03
0.154kO.005
0.70+0.02
0.192kO.005
0.84a0.02
O.l32t-0.004
0.84kO.02
0.134kO.008
0.81kO.03
0.083+0.003
l.Ol+O.OO
0.105kO.005
0.99kO.04
0.068kO.004
1.09kO.10
0.0531kO.003
1.00+0.04
0.033+0.001
0.97kO.08
0.036kO.002
1.02kO.03
0.044+0.002
1.07+0.04
0.047+0.002
0.88kO.03
A%
AdI%
-8,741
447
158
42
63
83
74
440
263
-1,409
568
535
1,710
-1,663
175
1,
-0.0001
0.002
0.006
0.024
0.016
0.012
0.014
0.002
0.004
-0.001
0.002
0.002
0.001
-0.001
0.006
9,
-33
412
29
18
41
28
25
-307
298
-25
374
42
-78
-23
14
TP
-0.030
0.002
0.035
0.055
0.025
0.035
0.040
-0.003
0.003
-0.040
0.003
0.024
-0.013
-0.043
0.073
*‘p
pill
-0.29
-5.5
5.2 '
0.40
7
13
21
32
12
25
12
21
13
23
-0.66
5.2
0.86
8.3
-1.8
-6.6
0.34
4.1
4.3
1.9
-1.1
1.4
-4.5
-1.5
8.4
12
JTh
c
%
3
2-.
2
Y
-I.
Q
2
s
g
h
?
196
Coale and B&and
yj200
E
-u
‘v, 100
AT
0
0.1
0.2
0.3
0.4
Dissolved
0.5
0.6
0.7
0.8
0.9
2341-h: 238U
Fig. 3 Dissolved 234Th residence time as a function
of 234Th 238U. Stippled area represents error estimates
about 74 and is based upon an uncertainty in 234Th
activity of 3.5% (average uncertainty, 1 SD for VERTEX 3).
In our estimates of 238U activities, we assigned an error of the order of 1% (1 SD) to
the salinity determination (in which we have
much greater confidence) and propagated it
along with all other errors. This error is small
with respect to the final uncertainties in our
determinations.
The average analytical
errors (SD) for VERTEX 2 dissolved and
particulate 234Th analyses were 6.7 and 7.5%.
For VERTEX 3, they averaged 3.5 and 3.4%.
For VERTEX 4, they averaged 4.6 and 5.9%.
The time scales over which 234Th: 238U disequilibria can be used is limited by the accuracy and precision with which the 234Th :
238U activity ratio can be determined. As
an example, Fig. 3 depicts dissolved residence time, r& as a function of the dissolved
234Th : 238U
activity
ratio; a 3.5% error
(VERTEX 3) is incorporated in this calculation. This error has been used to construct
error estimates about Td and is depicted as
the hatched area in Fig. 3. As can be seen
from this figure, the closer the system is to
equilibrium,
the more difficult it becomes
to evaluate residence times reliably. For example a dissolved 234Th : 238U activity ratio
of 0.90+0.03 yields Td = 3 113d, but a range
(+ 1 SD) of 223-462 d.
Measurements of light transmission using
deck and submarine photometers at VERTEX 2, 3, and 4 indicate 0. I % light levels
occur at about 100, 100, and 150 m, re-
spectively
(Broenkow
and Krenz 1982;
Broenkow et al. 1983; Broenkow and Reaves
1985). Temperature and nitrate profiles for
VERTEX 2, 3, and 4 indicate mixed-layer
thicknesses for these stations of 40, 45, and
25 m (Fig. 2). The shallow VERTEX 4 mixed
layer of 25 m is typical of summer stratification, whereas the deeper mixed layers at
VERTEX 2 and 3 are characteristic of autumn. In all cases, the euphotic zone extended deeper than the mixed layer, allowing slight illumination
of nutrient-rich
waters below.
Primary production at VERTEX 2,3, and
4 is relatively uniform throughout the mixed
layer and, in general, decreases monotonically with depth (Knauer et al. 1984). In
sediment trap collections from the 0- to
40-m layer, no resolvable detrital flux was
observed. However, a significant flux was
observed in sediment traps beneath the euphotic zone for all stations. New production
as a percentage of the 40-100-m primary
production at the VERTEX 2 site, for example, was 15% in terms of carbon-a percentage slightly less than those in eutrophic
regions (Knauer et al. 1984; Small and
Knauer pers. comm.). Furthermore, Small
and Knauer (pers. comm.) found that fecal
pellets from the more abundant, large (200-2,000 pm) microcrustaceans in the lower
layer at the VERTEX 2 location represented
slightly over 4% of the primary carbon production in the lower layer, but 30% of the
trap-estimated carbon flux from this layer.
Because the total particulate carbon flux out
of the 0-40-m layer was immeasurably small
and no fecal pellets were seen in the 30-m
trap, the fraction due to fecal pellet flux was
undefined. Those pellets released by zooplankton of any size in the upper layer must
have been fragmented and recycled there.
Discussion
Important
conclusions and hypotheses
can be drawn from these observations. First,
the upper layer is tightly coupled. Essentially all carbon and nitrogen fixed in this
layer is consumed and recycled within it.
Second, new production
is a prominent
component of primary production in the
lower layer. Third, due to the sharp nutri-
Stratified Th : U disequilibria
Cline/pycnocline,
new production is a negligible contribution to production in the upper layer. Consequently, the euphotic zone
consists of an upper, truly oligotrophic layer, overlying a lower layer with more eutrophic characteristics.
Our 234Th: 238U data, in general, are consistent with these observations and can be
used to further expand the concept of stratified euphotic systems with respect to elemental scavenging.
Dissolved scavenging- 234Th : 238U disequilibria integrated over the depth of the
mixed layer yield dissolved 234Th residence
times of 52, 62, and >320 d for VERTEX
2, 3, and 4. These observations are consistent with other calculations of oligotrophic
mixed-layer residence times (Bruland and
Coale in press). However, in the region immediately below the mixed layer, residence
times for dissolved 234Th drop precipitously
to minima of 8,25, and 42 d for VERTEX
2, 3, and 4. These correspond to scavenging
rate constants @d of 0.13, 0.040, and 0.024
d-l, which are characteristic of the coastal
and eutrophic scavenging rate constants associated with high production
regimes
(Coale and Bruland 1985). Moreover, these
results indicate that net scavenging rates are
more closely related to new production
rather than to total production.
It is important to make a distinction here
between net and gross scavenging. The combination of high productivity
and negligible
particle flux for the mixed layer implies a
high degree of in situ recycling within this
layer. Recycling processes are a result of
metazoan and microbial oxidative decomposition and dissolution of particulate phases. It is possible, and even likely, that gross
scavenging of 234Th in the upper layer is
considerably higher than in the lower layer,
but metazoan and microbial particle destruction effectively recycles this particulate
Th back into solution so that a low net removal is observed.
The net flux of 234Th from the dissolved
to particulate phase (J=h) is higher in the
lower euphotic layer than in the upper layer,
giving rise, in every case, to the pronounced
minimum
in dissolved phase 234Th : 238U
ratios in the lower layer (Fig. 2). This min-
197
imum coincides with greater particle abundance in the lower layer, greater total particle flux out of the lower layer, and greater
partitioning of the particle flux of the lower
layer into fecal pellet flux.
Within the mixed layer, all parameters
except light are relatively
uniform
and
evenly distributed about a depth-integrated
mean. However in the lower layer there are
gradients of nitrate, light, temperature, and
density so that an integrated mean of 234Th
is no longer representative of scavenging
throughout this layer. The lower layer is
highly stratified. Complex ecological and
biogeochemical
processes are structured
over this gradient. The processes that control 234Th scavenging within this region are
complex and can lead to large variations in
scavenging rate (Table 1). In every case the
minimum in dissolved 234Th activities, the
minimum in dissolved 234Th residence time,
and the maximum in scavenging rate and
flux for the conversion of dissolved to particulate phases corresponds closely with the
depth of the subsurface pigment maximum
(Fig. 2). A subsurface pigment maximum in
the euphotic zone is a well documented and
prominent feature of both Pacific and Atlantic oligotrophic
regimes (e.g. Hayward
and McGowan 1985; Takahashi and Hori
1984; Anderson 1969; Platt et al. 1983). We
believe that the bulk of the new production
and elemental scavenging are fundamentally linked in some way to this pigment
maximum.
Particle removal- In every case we observe greater integrated removal of particulate 234Th (PTh) in the lower layer than in
the upper layer (Table 1). This finding is
also consistent with the observations that
particulate flux out of the euphotic zone occurs principally from the lower layer (Knauer
et al. 1984; Small and Knauer pers. comm.).
With the exception of VERTEX 4, particulate 234Th residence times are longer in the
lower layer. These longer particle residence
times in the lower layer during VERTEX 2
and 3 seem contrary to the fact that particulate flux is greater from the lower layer
than from the upper layer. However, particulate residence time is a function of removal rate (PTh) as well as the activity, or
Coale and B&and
198
Table 2. Sediment trap and model-derived
mates of 234Th flux (dpm m-2 d-l).
----
ML.ML*
trap sample
or depth integral
Particulate
234Th flux
VERTEX 2
MLML 30-m trap
30-m integrated PTh
MLML 120-m trap
120-m integrated P,.,,
835+45-t
570
1,878+48
1,222
VERTEX 3
MLML 80-m trap
80-m integrated PTh
MLML 120-m trap
120-m in tegratcd PTh
1,505*50t
1,600
73Ok 14
1,861
VERTEX 4
MLML 150-m trap
150-m integrated PTh
esti-
671
630
* Moss Landing Marine Laboratories.
j’ Samples in which large numbers of killed zooplankton
were not completely removed before analysis; therefore they arc thought to be artifactually
high in 234Th. Trap deployment
odepths do not necessarily
correspond
to depths of upper and lower euphotic zones for these
slatlons.
size, of the particulate reservoir. Coale and
Bruland ( 198 5) suggested that particulate
removal is controlled by the types and
abundances of grazing zooplankton present.
At the VERTEX 2 and 3 sites, Small and
Knauer (pers. comm.) observed greater
numbers of large zooplankton in the lower
euphotic layers than in the upper layers.
These observations suggest that particulate
234Th residence times should be shorter in
the lower layer, not longer. Particulate 234Th
residence times at the VERTEX 2 and 3
stations seem to vary mainly as a function
of the size of the particulate 234Th reservoir.
In all cases the activity of particulate 234Th
is greater in the lower layer and coincident
with a large pigment maximum; this is also
a region in which greater suspended particulate concentrations were observed.
The POC maximum was not necessarily
coincident with the pigment maximum at
the stations for which we have POC data
(VERTEX 2 and 4). For VERTEX 2, integrated mixed-layer and lower-layer POC
values were 44 and 45 pg C liter - I, respectively, with a maximum of 64 pg C liter-’
at 60 m. For VERTEX 4, integrated mixedand lower-layer POC values were 30 and 29
pg C liter-l, respectively, with a maximum
of 53 pg C liter-l at 50 m (G. Knauer et al.
pers. comm.). Pigment maxima for VERTEX 2 and 4 occurred at 50 and 90 m. In
both cases the 234Th scavenging rate is greatest in or near the pigment maximum and is
not necessarily associated with the particle
maximum.
The VERTEX 2 and 3 stations both show
a lower, but significant, flux of 234Th from
the upper layer via particle removal (PTh).
This finding is somewhat contrary to the
sediment trap and fecal pellet observations
of Knauer et al. (1984) and the fact that
some type of upper-layer particle removal
process is required to both establish the twolayer system initially and to maintain P*h
as observed with water-column
234Th disequilibria. This apparent discrepancy needs
to be reconciled. Knauer et al. (1984) and
Small and Knauer (pers. comm.) could resolve no fecal pellet flux from the 0- to 40m depth interval; however, a large number
of dead zooplankton (swimmers, killed by
the trap preservative) were observed in the
traps. Killed zooplankton were not quantitatively removed in samples from particle
interceptor traps (PITS) analyzed for 234Th;
thus, the 234Th flux values could have been
artifactual. Sediment trap results and modcl-derived estimates of 234Th flux for VERTEX 2,3, and 4 are given in Table 2. During
VERTEX 3, the shallowest trap (80 m) was
set 35 m below the mixed layer so we have
no direct evidence to account for the apparent discrepancy in water-column-derived PTh and trap-measured P,,. We can
suggest, however, that particulate 234Th may
be transported out of this layer by large,
vertically migrating zooplankton. The latter
scenario would serve to short-circuit
particle transport from this layer.
Some evidence supports this active transport scenario. The pelagic red crab Pleuroncodes planipes was a dominant component of the meroplankton during VERTEX
2, having been observed in net tows in the
mixed layer and the lower euphotic zone.
Pleuroncodes planipes is known to inhabit
waters in this area from the surface to 300
m. The fecal pellets of this organism were
conspicuous components of the trap flux at
120 m but were absent from the 30-m trap.
If micronectonic organisms such as P. plan-
StratiJied Th : U disequilibria
ipes were feeding in the upper layer and defecating at depth, it would help to account
for the apparent discrepancy between watercolumn and PIT measurements of PTh.
Pleuroncodes planipes was not abundant
during VERTEX 3: therefore, the vertical
migration scenario must remain tentative.
Other vertical migrators, such as euphausiids, were collected at all VERTEX stations, including VERTEX 4, but their relative abundances are not known, and their
possible roles in vertical transport of material via vertical migration also are unknown.
For VERTEX 4, where the 25-m sample
shows total 234Th in slight excess of 238U, a
migratory active transport of materials from
the upper layer need not be evoked. We
consider the mixed layer of VERTEX 4 to
be in near equilibrium,
the epitome of an
oligotrophic system. Here PTh is essentially
zero and both Td and r, are very long with
respect to nonradioactive
removal. These
values represent a very tightly coupled upper layer overlying a layer of moderate new
production.
Conclusions
234Th: 238U disequilibrium
data can be
used to elucidate and support the concept
of stratified oceanic euphotic zones. Moreover, 234Th: 238U disequilibria
can be used
to describe oceanic elemental scavenging and
particle transport rates within these stratified systems. Disequilibrium
data indicate
that net rates of particle reactive elemental
transfer from the dissolved to particulate
phase are low in the oligotrophic mixed layer. Scavenging and particle transport rates
are highest in or near the pigment maximum associated with the lower euphotic
layer of the two-layer system. The rates then
generally decrease to aphotic midwater values. Gross rates of elemental scavenging
have yet to be assessed. Differences in scavenging between upper and lower layers suggest that net rates of elemental scavenging
vary as a function of new production rather
than total production.
Stratified euphotic
zones appear to structure elemental scavenging and particle transport and may be a
feature of oligotrophic
euphotic zones in
199
general. Data for this study reflect only summer and autumn conditions. Due to the seasonally variable nature of stratification, we
would expect to see seasonally variable
scavenging and particle flux from these layers as well.
References
ANDERSON,G. C. 1969. Subsurfacechlorophyll
maximum in the Northeast Pacific Ocean. Limnol.
Oceanogr. 14: 386-39 1.
BACON,M. P., AND R. F. ANDERSON. 1982. Distributions of thorium isotopes between dissolved and
particulate forms in the deep sea. J. Geophys. Res.
87: 2045-2056.
BHAT, S: G., S. KRISHNASWAMI,D. LAL, RAMA, AND
W. S. MOORE. 1969. 234Th/238U ratios in the
ocean. Earth Planet. Sci. Lett. 5: 483-491.
BROENKOW,W. W., AND R. KRENZ. 1982. Oceanographic results from the VERTEX-II
particle trap
experiments off Manzanillo,
Mexico, 26 October
to 18 November, 198 1. Moss Landing Mar. Lab.
Tech. Publ. 82-l.
-,
A. J. LEWITUS,AND R. E. REAVES. 1983.
Oceanographic results from the VERTEX-3 particle interceptor trap experiment off Central Mexico, October-December,
1982. Moss Landing Mar.
Lab. Tech. Publ. 83-l.
AND R. E. REAVES. 1985. Oceanographic resuits from the VERTEX-4
particle-trap
experiment north of Hawaii July-August
1983. Moss
Landing Mar. Lab. Tech. Publ. 85-I.
BRULAND,K. W., AND K. H. COALE. In press. Surface
water 234Th/238U disequilibria:
Spatial and temporal variations of scavenging rates within the Pacific Ocean. Zn Dynamic processes in the chemistry of the upper ocean. NATO Conf. Ser. 4. Mar.
Sci. Plenum.
COALE, K. H., AND K. W. BRULAND. 1985. 234Th:
238U disequilibria
within the California Current.
Limnol. Oceanogr. 30: 22-33.
DUGDALE,R. C. 1967. Nutrient limitation in the sea:
Dynamics, identification
and significance. Limnol. Oceanogr. 12: 685-695.
AND J. J. GOERING. 1967. Uptake of new and
regenerated forms of nitrogen in primary productivity. Limnol. Oceanogr. 12: 196-206.
EGGIMANN,D. W., AND P. R. BETZER. 1976. Decomposition and analysis of refractory oceanic suspended materials. Anal. Chem. 48: 886-890.
EPPLEY,R. W., ANDB. J. PETERSON.1979. Particulate
organic matter flux and planktonic new production in the deep ocean. Nature 282: 677-680.
HARRISON,W. G., D. DOUGLAS, P. FALKOWSKI,G.
ROWE, AND J. VIDAL. 1983. Summer nutrient
dynamics of the Middle Atlantic Bight: Nitrogen
uptake and regeneration. J. Plankton Res. 5: 539556.
HAYWARD,T. L., AND J. A. MCGOWAN. 1985. Spatial
patterns of chlorophyll, primary production, mat- .
roplankton biomass and physical structure in the
200
Coale and B&and
central North Pacific Ocean. J. Plankton Res. 7:
147-167.
HERBLAND, A., A. LEBOUTEILLER, AND P. RAIMBAULT.
198 5. Size structure of phytoplankton
biomass in
the equatorial Atlantic Ocean. Deep-Sea Res. 32:
8 19-836.
JENKINS, W. J., AND J. C. GOLDMAN. 1985. Seasonal
oxygen cycling and primary production in the Sargasso Sea. J. Mar. Res. 43: 465-49 1.
KAUFMAN, A., Y.-H. LI, AND K. K. TUREKIAN. 198 1.
The removal rates of 234Th and 228Th from waters
of the New York Bight. Earth Planet. Sci. Lett. 54:
385-392.
KNAUER, G. A., AND J. H. MARTIN. 198 1. Primary
production and carbon and nitrogen fluxes in the
upper 1,500 m of the northeast Pacific. Limnol.
Oceanogr. 26: 181-182.
_AND D. M. KARL. 1984. The flux of
pa&iculate’organic matter out of the euphotic zone,
p. 136-150. In Global ocean flux study. Proc.
Workshop Natl. Acad. Sci.
KNAUSS, K. G., T. Ku, AND W. S. MOORE. 1978. Radium and thorium isotopes in the surface waters
of the East Pacific and coastal southern California.
Earth Planet. Sci. Lett. 39: 235-249.
Ku, T.-H., K. G. KNAUSS, AND G. G. MATHEIU. 1977.
Uranium in the open ocean: Concentration
and
isotopic composition. Deep-Se:a Res. 24: 13-26.
MCKEE, B. A., D. J. DEMASTER, AND C. A. NITTROUER.
1984. The use of 234Th/238U disequilibrium
to
examine the fate of particle-reactive
species on the
Yangtze continental shelf. Earth Planet. Sci. Lett.
68: 43 l-442.
MATSUMOTO, E. 1975. 234Th-238U radioactive
disequilibrium in the surface layer of the ocean. Geochim. Cosmochim. Acta 39: 205-212.
PLATT, T., D. V. SUBBA RAo, AND B. IRWIN. 1983.
Photosynthesis of picoplankton in the oligotrophic
ocean. Nature 301: 702-704.
SANTSCHI, P. H., Y.-H. LI, AND J. BELL. 1979. Natural
radionuclides
in the water of Narragansett Bay.
Earth Planet. Sci. Lett. 45: 201-213.
STEELE,J. H., AND C. S. YENTSCH. 1960. The vertical
distribution
of chlorophyll.
J. Mar. Biol. Assoc.
U.K. 39: 2 17-226.
TAKAHASHI, M., AND T. HORI. 1984. Abundance of
picoplankton in the subsurface chlorophyll maximum layer in subtropical and tropical layers. Mar.
Biol. 79: 177-186.
TUREIUAN, K. K., AND L. H. CHAN. 197 1. The marine
geochemistry of the uranium isotopes, 230Th and
231Pa, p. 3 1 l-320. In Activation analysis in geochemistry and cosmochem is try. Universitetsforlaget.
VENRICK, E. L. 1982. Phytoplankton
in an oligotrophic ocean: Observations
and questions. Ecol.
Monogr. 52: 129-l 54.
Submitted: 4 September 1985
Accepted: 5 August 1986