MEDFLUX: RELATIONSHIPS AMONG BALLAST, PARTICULATE ORGANIC CARBON AND
234
TH ACTIVITIES AND FLUXES AT THE DYFAMED SITE, NORTHWESTERN MEDITERRANEAN
J. K. Cochran
1*
, J. C. Miquel
2
, S. W. Fowler
2
, B. Gasser
2
, J. Szlosek
1
, A. Rodriguez-y Baena
2
, D. Hirschberg
1
,
G. Stewart
1
and P. Masqué
3
1
Marine Sciences Research Center, Stony Brook University, Stony Brook, New York 11794-5000 USA
* kcochran@notes.cc.sunysb.edu
2
IAEA Marine Environment Laboratory, 4 Quai Antoine 1er, MC98000 Monaco
3
Universitat Autonoma de Barcelona, Bellaterra, Spain
Abstract
Standard approaches for measuring the distribution and ?ux of
234
Th (in situ pumps, sediment trap) have been applied at the DYFAMED
site, NW Mediterranean and complemented by analysis of samples from traps that collect and separate particles into fractions according
to sinking velocity, and particles collected in a “net trap” and separated into sinking velocity using an elutriator on board the ship. The
results show that >70% of the sinking ?ux of
234
Th is carried by particles sinking at velocities greater than ~100 m d
-1
. Measurements of
particulate organic carbon and elements corresponding to mineral ballast in the sinking particles will help to better understand the
relationships among these parameters.
Rapp. Comm. int. Mer Médit., 37,2004
181
Sinking ?uxes of particulate organic carbon (POC) in the oceanic
water column are frequently measured using the disequilibrium
between the natural radionuclide
234
Th and its parent
238
U. Such
estimates require measurement of both the extent of disequilibrium
and the POC/
234
Th ratio on the sinking particles. The latter is
commonly estimated through measurement of the POC/
234
Th ratio in
a large (>50 to 70
µ
m) filterable fraction or in sediment trap material.
The relationship between the POC/Th in the filterable and sinking
particles remains uncertain, however. As part of the MedFlux
program, we used in situpumps to measure the patterns of Th/U
disequilibrium in the water column at the French JGOFS time-series
DYFAMED site in the northwestern Mediterranean. In addition, IRS
swimmer-exclusion sediment traps were deployed at 200 m to collect
both a regular time series of particle ?ux and a composite sample with
individual cups programmed to collect particles separated according
to sinking velocity. A large sample of particles sinking through 200
m was collected by means of a newly-designed net trap deployed for
3 days, and subsequently separated into sinking velocity fractions
using an elutriator.
Several water column profiles of particulate and dissolved
234
Th
were collected over a one-week period in May 2003 (Fig. 1). The data
show a pronounced minimum in total
234
Th (particulate + dissolved)
centered on the depth of the chlorophyll maximum. Fluxes of
234
Th
at 200 m (calculated applying a steady state model to the
234
Th
deficit) were 2000 – 3000 dpm m
-2
d
-1
. These ?uxes were in the range
of those observed in a time-series sediment trap deployed at 200 m
during the previous 60 days, although they more closely corresponded
to trap ?uxes measured early in the period (Fig. 2).
Fig.1 Water column profiles of
234
Th (particulate:circles; dissolved:tri-
angles and total:squares) and
238
U (diamonds).
Fig.2 Time series of
234
Th ?uxes at 200 m.
Particles separated by sinking velocity using both the sediment trap
and net trap/elutriator methods showed similar results. More than
50% of the
234
Th ?ux in the integrated sediment trap sample was
carried in particles sinking faster than 196 m d
-1
(Fig. 3). In the net
trap sample collected May 6-8, more than 70% of the
234
Th collected
was present on particles sinking at rates >100 m d
-1
. The specific
activity of
234
Th showed no clear relationship to sinking velocity.
Future measurements of POC and ballast elements (Si, Ca, Al) will
clarify the relationships between Th ?ux and particle composition.
Fig.3. Integrated
234
Th ?ux vs particle sinking velocity.
