Primaryinformation
production on
andC-sequestration
potential for carbon
export
in
Gaining
efficiency
using
naturally
iron-fertilized
in the Southern
Ocean
a production
/ export / waters
remineralisation
toolbox:
the
S.O. naturally Fe-fertilized areas study-case
Anne-Julie Cavagna
Frank Dehairs
Stéphanie H.M. Jacquet
Frédéric Planchon
Antarctic Session
Natural Fe-fertilized open ocean zones in the S.O.
Constraint of blooms by circulation & topography
SeaWiFS chl-a images in October and December 1998 (from Pollard et al., 2007)
What do we learn from
comparative study of Fe-replete
/ Fe-deplete areas & time series
located in FeNX sites ?
CROZEX
leg 1 (Nov. 2004-mid-Dec. 2004)
leg 2 (mid-Dec. 2004-Jan. 2005)
KEOPS
leg 1 (Jan.-Feb. 2005) SUMMER
leg 2 (Oct.-Nov. 2011) SPRING
SAZ-Sense
Jan.-Feb. 2007
SUMMER
CROZEX (Spring – early Summer 2004/05)
North area
LARGE LONG EARLY BLOOM
High surface chl-a
high productivity zone
Defined as “bloom / Fe-replete”
N
S
Surface Chl a (mg m-3)
South area
SMALL SHORT LATE BLOOM
Low surface chl-a
low productivity zone
Defined as “HNLC control / Fe-deplete”
CROZEX (Spring – early Summer 2004/05)
Morris and Sanders, 2011 (GBC)
Seasonal integration
Hide shorter timescale events
- Significant increased level of integrated PP in the N. compared to the S.
-- shallow seasonally integrated export, annually integrated deep water POC flux
and core-top organic carbon accumulation enhanced 2 to 3 fold as a result of the
iron-fertilized bloom (Pollard et al., 2009 - Nature)
CROZEX (Spring – early Summer 2004/05)
Leg 1
N-S gradient
Leg 2
No N-S gradient seen once
the modest bloom occurred
in the south
≈ 180 mgC m-2 d-1
234Th
Post-bloom EP insensitive
to size of bloom
≈ 60 mgC m-2 d-1
Nov. => mid-Dec.
derived export rate:
Mid-Dec.. => Jan.
Morris et al., (2007) DSR2
Why similar export in high productive & low productive zone during Leg 2 ?
 North = High Biomass Low Export zone ? (HBLE – Lam & Bishop, 2007 DSR II)
 Miss the high export rate at bloom peak ?
 New and export production are not equivalent, with this lack of equivalence
being particularly pronounced in the north (Fe-replete area)
The toolbox – production / export / remineralisation
Fe, nutrients, light, stratification
POC (µM)
0m
Net primary
production
NetPP
(mgC m-2 d-1)
0.5
1% 5%
10%
20%
Remin.
100 m 40%
30%
EP (mgC m-2 d-1)
Export
Remin. MR (mgC m-2 d-1)
0.4
EP/Net PP
1000 m
Export
0.3
Remin.
POC attenuation curve
0.2
0.1
Carbon sequestration efficiency
(deep carbon export relative to surface netPP)
0.0
0.0
0.2
0.4
0.6
EP700/EP = 1 - MR/EP
0.8
1.0
Based on the idea of
Buesseler & Boyd L&O (2009)
SAZ-Sense (Summer period 2007)
STF
2.0
ZC
P3
EAC
SAZ-N
1.6
-1
chl a surface (µg l )
STZ
P1 #3
1.2
P3
P1
P1 #2
0.8
SAF-N
SAF-S
P2
SAZ-S
0.4
3 repeat measurement /
station in 1 week
P1 #3
0.0
0
25
50
75
100
125
-2
150
PFZ
AZ
P2
-1
integrated GPP (mmolC m d ) -euphotic layerSurface Chl-a (mg m-3)
P1 = 929 ± 808 mgC m-2 d-1
=> 70 mgC m-2 d-1
P2 = 424 ± 18 mgC m-2 d-1
=> 32 mgC m-2 d-1
P3 = 680 ± 96 mgC m-2 d-1
=> 5.4 mgC m-2 d-1
P3
P1
P2
Export 100 m vs. production EP100/GPP
SAZ-Sense (Summer period 2007)
1% 5% 10%
20%
30%
0.5
0.4
0.3
P1
0.2
P2
P3
0.1
0.0
0.0
0.2
0.4
0.6
0.8
1.0
Export 600 m vs. export 100 m T600 = EP600/EP100
P3 => High Biomass Low Sequestration system ?
Stable system less efficient than versatile system for carbon export + sequestration
KEOPS (KEOPS 1 Summer period - 2005)
A3 site
INSIDE THE BLOOM
High surface chl-a
high productivity zone
Defined as “bloom / Fe-replete”
C11 site
OUTSIDE THE BLOOM
Low surface chl-a
low productivity zone
Defined as “HNLC control / Fe-deplete”
KEOPS (KEOPS 1 Summer period - 2005)
Highly active bacterial community
On-shelf
Prevalence of regenerated production
and low uptake of NO3 above the
Plateau  proportionally low export.
Plateau surface waters operate as a
High Biomass Low Export system, but
since subsurface remineralisation is
relatively limited there still is an
important fraction of C left for deep
sequestration. However overall the offshelf system appears as the most
efficient site for C sequestration
15.2%
28.3%
KEOPS (KEOPS 2 Spring period - 2011)
11 novembre 2011
E5
E4W
Courtesy from Y-H. Park
(MODIS Chl-a biomass +
data from surface buoy
and altimetry (Nov. 2011)
E1
F
E4E
E3
R
A3
From expedition & first workshop data analysis: 3 clusters + reference station:
 Reference station (HNLC and low Fe) : R
 Cluster 1 (productive sites south of PF) : A3-2 and E4W
 Cluster 2 (stationary permanent meander south of PF): E stations: E1, E3, E4E, E5
 Cluster 3 (productive site on to north of the Polar Front): NPF
Toolbox data KEOPS 1 & KEOPS 2
C-export production
Meso-remin.
Net PP
234Th proxy
Particulate Baxs proxy
(mgC m-2 d-1)
-2
-1
(mgC m d )
(mgC m-2 d-1)
C-sequestration
Efficiency
(mgC m-2 d-1)
R
132 ± 22
23 ± 17
86.3
To be investigated
C11 (summer)
300
120
36
85
NPF
3380 ± 145
53 ± 07
41.2
16.9
E4W
3287 ± 83
87 ± 12
65.5
32.9
A3-2 (spring)
2172 ± 230
47 ± 22
17.6
21.7
A3 (summer)
1460
250
28
222
E1 (day 0)
578 ± 54
156 ± 18
42.0
115.6
E3 (day 5)
748 ± 103
159 ± 16
32.3
130
E4E (day 14)
1037 ± 130
On going
57.4
On going
E5 (day 20)
1064 ± 126
99 ± 11
62.0
74.5
• 234Th derived integrated export below 100m exceeds 200m trap C-export (T. Trull pers.
communic.) by 20 to 60%
Toolbox data KEOPS 2
NPF
New production / Export production
3000
2500
A3-2
2000
1500
New prod/Net prod
Exp.prod/Net prod.
1000
E1
500
E4EE5
E3
E4W
R
0
0
1000
2000
3000
4000
Net primary production
In accordance with CROZEX (Morris et al., 2007 – DSR2), we observe for KEOPS 2 an
evidence for a decoupling of new and export production. With also the effect being
most apparent in the high productive area (for CROZEX the effect was most
apparent within the northern bloom area)
KEOPS Integrated Information
1%
0.5
5%
10%
20%
30%
40%
A3
C11 (K1 HNLC)
C11
Meander
EddyE
NPF
ThE = EP/NetPP
EP/NP
0.4
0.3
E1 (day 0)
Spring
0.2
A3-2
E3 (day 5)
A3 (K1)
E5 (day 20)
Summer
0.1
E4W
NPF
0.0
RK2 (50°S-66°E)
Early spring
0.0
A3-2
0.2
0.4
0.6
EP700/EP
EP700/EP
= 1 – MR/EP
KEOPS 2 (spring period) and KEOPS 1 (summer period)
 High surface chl-a sites = high production – low sequestration
 Meander E & A3 site at keops 1 and 2 = highlight a seasonal cycle
0.8
1.0
KEY-POINTS
Deep carbon sequestration efficiency is related to the type of production regime
Low Biomass systems (E stations at K2 in early season; C11 at K1) seem to be more
efficient in terms of C-sequestration than High Biomass systems (K2: E5 cluster 1 and 3;
K1: A3)
** High Biomass Low Sequestration vs. Low Biomass High Sequestration **
=> Not in contradiction with Fe-replete areas exporting more than Fe-deplete areas
Example for K1: PP at C11 (Fe-deplete area / HNLC) is only 20% of PP at A3 (Fe-replete
area) => C11 sequestration = 38% A3 sequestration
Is there evidence for a temporal succession from LBHS to HBLS over the season ?
LBHS at the early stage of the productive season
Rapid transition to HBLS was ongoing for E stations, while clusters 1 and 3 were
already HBLS at the start of the study
=>At the end of the season HBLS conditions (A3 Keops2) returned to LBHS (A3 Keops1)
QUESTION :
Do systems keep the ‘LBHS’ status during winter ? What is the strength of the
biological pump in winter?
Putting the pieces together
3 / Primary production & potential for carbon export
17
Natural Fe-availability and enhanced surface Chl a does not always
reflect enhanced integrated production and deep carbon export
Fe, nutrients, light, stratification
POC (µM)
0m
Gross primary
production
remineralization
100 m
600 m
Export
remineralization
Export
remineralization
POC attenuation curve
different systems can have the
same deep export efficiency
What do we learn from previous FeNXs inter-comparison ?
9
Natural Fe-availability and enhanced surface Chl a does not always reflect
enhanced integrated production and deep carbon export, especially at the end
of the productive season
different systems can have the
same export export efficiency
and inversely
Fe, nutrients, light, stratification
0m
100 m
POC (µM)
Gross
primary
production
remineralization
End of the productive season,
naturally Fe-fertilized sites seems
to function as HBLE systems
=> Needs further investigations
Export
remineralization
600 m
These 2 studies occurred at the
end of the productive season
Export
remineralization
POC attenuation curve
Key observations
12
High surface productivity in the Kerguelen Islands area is perhaps not only due to natural iron fertilization but also to
vicinity with Polar Front (mesoscale frontal dynamics boost primary production- Strass et al. 2002 – DSR II)
If nutrient consumption efficiency is increased by iron artificial addition, what will remain for the low latitude regions
nutriently supplied by Antarctic Intermediate Water (Sarmiento et al., 2004 - Nature) ?
Tamburini et al. (2009 –DSR II) demonstrate from 200 to 1500m that pressure decrease the number of prokaryotes
attached to aprticles and the apparent activity of free-living prokaryotes. This helps to explain why fast sinking particles
such as fecal pellets, but possibly also including fast sinking marine snow aggregates, can fall through the water column
with minimal degradation.
Looking on A station, we join one of the De Brauwère et al. 2013 conclusion being that increasing analytical information
throughout the duration of the bloom would strongly help to upgrade and tune models
Deep carbon export efficiency using the proposed toolbox is an encouraging way to gain information on the biological
carbon pump. The important point is to carefully take MLD and EZD into account in order to avoid dangerous
misestimation.
KEOPS 2: Raw information is available to mature the 3 flux estimation needed to obtain the relative global view of
studied systems
R station shows a peculiar functioning: leads to the question of winter primary production
Preliminary results. Have to be carefully validated together (depth layers).
The toolbox – production / export / remineralisation
13C-assimilation (Net
PP) and 15NO3 / 15NH4-uptake rates (f-ratio – New production)
 Euphotic zone depth integrated parameters (7 depths measurements between 75 and 0%
light attenuation)
 24 h incubation experiments (daily Net PP = Gross PP + C-loss)
 15N-NO3- dilution experiment to measure nitrification in the euphotic zone
Carbon export below the surface water
using ISP sampling 234Th proxy

Fe, nutrients, light, stratification
POC (µM)
234Th
deficit / excess depth profile
measurement
 C-export conversion using C/234Th ratio in
particle (2 size classes at each sampling
depth)
0m
Net primary
production
Remin.
100 m
(See Savoye et al. 2008 DSR2)
Export
Mesopelagic carbon remineralisation
using particulate Baxs proxy
Remin.
700 m
 Baxs ICP-MS measurements
 Dehairs et al. (1997) DSRII algorithm to convert
Baxs content into final POC mineralization rate
(S.H.M. Jacquet – poster 361)
Export
Remin.
POC attenuation curve
KEOPS
Isotopic model of oceanic silicon
cycling: the Kerguelen Plateau
case study (de Brauwere et al., in
revision for DSR I)
 Having
additional measurement
during the season would tremendously
help to constrain the bloom peak and
hence the rate parameters
 A puzzling result of this modeling
exercise is that seasonally-integrated
Si-uptake flux above the plateau is
lower than off the plateau while it might
be expected that above the plateau
more production occur due to the
fertilization effect.
CROZEX
Natural Fe-fertilized open ocean zones in the S.O.
S.O. species have overcome the
antagonistic iron-light relationship by
increasing size rather than number of
photosynthetic
units
under
low
irradiance resulting in an acclimatation
strategy that does not increase their
cellular iron requirement.
xx
Planchon et al. 2013 BGH transect (summer period from South Africa to
northern Weddell gyre) => same range than Exp. Prod. at KEOPS 2 R station
C-flux at 100m (SS model)
(mgC m-2 d-1)
21,6
10,8
20,4
27,6
31,2
39,6
42,0
56,4
61,2
51,6
39,6
Regime of production – surface water (euphotic zone)
Net PP
(mgC m-2 d-1)
f-ratio
Exportable prod. Euphotic Zone
U-NO3/(U-NH4+U-NO3) Net PP x f-ratio
nitrification
R
132 ± 22
0.41
50
yes
NPF
3380 ± 145
0.81
2738
yes
E4W
3287 ± 83
-
-
no
A3-2
2172 ± 230
0.87
1890
Yes
E1 (day 0)
578 ± 54
0.70
405
no
E3 (day 5)
748 ± 103
0.59
441
yes
E4E (day 14)
1037 ± 130
0.67
695
no
E5 (day 20)
1064 ± 126
0.64
681
no
R station => control HNLC with low Net PP
A3 => KEOPS 2 = 181.0 ± 19.2 mmolC m-2 d-1 (f-ratio = 0.9) EARLY SPRING
KEOPS 1 = 80.6 ± 5.6 mmolC m-2 d-1 (f-ratio = 0.6) SUMMER
E stations => Effective temporal variation through 3 to 4 weeks monitoring
11
Carbon export – below the surface water (100m horizon)
Euphotic layer
C-export production
234Th proxy
ThE-ratio (%) depth (m)
Net PP
-2 -1
EP:NetPP 0.3% (0%) PAR
(mgC m-2 d-1) (mgC m d )
Mixed layer
depth (m)
R
132 ± 22
23 ± 17
17
116 (-)
=
107
NPF
3380 ± 145
53 ± 07
1.6
33 (52)
=
29
E4W
3287 ± 83
87 ± 12
03
42 (67)
<
57
A3-2
2172 ± 230
47 ± 22
02
49 (78)
<
163
E1 (day 0)
578 ± 54
156 ± 18
27
80 (126)
>
64
E3 (day 5)
748 ± 103
159 ± 16
21
86 (137)
>
27
E4E (day 14) 1037 ± 130
On going
On going
42 (67)
<
70
E5 (day 20) 1064 ± 126
99 ± 11
09
69 (110)
>
58
• Evidence for carbon export in pre-bloom conditions.
• 234Th derived integrated export below 100m exceeds 200m trap C-export (T. Trull pers.
communic.) by 20 to 60%
• K2 C-export fluxes (early spring) are generally smaller than during K1 (summer)
12
Remineralisation – mesopelagic zone (MLD - 700m)
Net PP
(mgC m-2 d-1)
C-export production
234Th proxy
(mgC m-2 d-1)
Meso-remin.
Particulate Baxs proxy
(mgC m-2 d-1)
Meso-remin:EP
(0<value<1)
R
132 ± 22
23 ± 17
86.3
3.75
NPF
3380 ± 145
53 ± 07
41.2
0.78
E4W
3287 ± 83
87 ± 12
65.5
0.75
A3-2
2172 ± 230
47 ± 22
17.6
0.37
E1 (day 0)
578 ± 54
156 ± 18
42.0
0.27
E3 (day 5)
748 ± 103
159 ± 16
32.3
0.20
E4E (day 14)
1037 ± 130
On going
57.4
On going
E5 (day 20)
1064 ± 126
99 ± 11
62.0
0.63
13
R : meso-remineralization strongly exceeds C-export below the euphotic zone / mixed layer.
 Though same magnitude of temporal scale integration for 234Th and Baxs proxies (several
weeks), EZ C-export & meso-remineralization seems to be decoupled.
 If ambient mesopelagic water are saturated in BaSo3, barytine cristals will not be
dissolved: to be checked.