The Nitrogen Isotopic Record From the Peru-Chile Suboxic Zone; Distinguishing Internal and External Signals Across the Last Deglaciation

swedishstreakMécanique

22 févr. 2014 (il y a 3 années et 4 mois)

47 vue(s)

The Nitrogen Isotopic Record From the Peru
-
Chile Suboxic Zone;
Distinguishing Internal

and
External Signals Across the Last Deglaciation

M
. A. Altabet
1
,
S.C.
Bova
2
,
T.
Herbert
2
,
Y.
Rosenthal
3
,
J. Kalansky
3



Acknowledgments:
Funding from the NSF P2C2
program, Jen
Larkum

and
Rehka

Singh for technical support

1) Introduction


The Peru
-
Chile suboxic zone is one of 3 major open ocean regions where vanishingly small subsurface O
2

concentrations enable
microbially
-
mediated fixed N loss. This loss associated with denitrification and anammox processes globally is a predominate
control on marine N cycling and oceanic N inventory. Past variations in the extent of low O
2

and N
-
loss are recorded in the δ
15
N of
underlying sediments of these regions and prior work has shown climate sensitivity on centennial to orbital time scales. Amon
gst

a
number of findings has been a sharp and early rise in δ
15
N at the beginning of the last deglaciation. Outstanding questions include
the nature of the forcing of this rapid deglacial increase in N loss and the relative contributions of imported vs. system ge
ner
ated
signals. Of the latter, changes in surface NO
3
-

utilization have the potential to also contribute to the sediment δ
15
N record.

#
B43C
-
0412

1
SMAST
, University of Massachusetts
Dartmouth

2
Geological Sciences, Brown
University

3
Institute
of Marine and Coastal Sciences, Rutgers University

2) Core Location and



Regional Context



To address these issues,
δ
15
N records
were constructed
from two long pistons cores recently collected on the
northern Peru margin in the vicinity of 4
°
S
(Fig. 1)
.
CDH
-
23
and CDH
-
26 were raised from 350 and 1000 m
depth.
These sites were chosen in expectation of a) high
continuous sedimentation rates from the LGM to present
and b) overlying water column characteristics are
representative of the primary source water for the Peru
-
Chile OMZ.


The
primary subsurface water mass constituting the Peru
-
Chile suboxic zone is sourced in the equatorial
undercurrent (‘13
°
C water’) and enters this system through
the Peru
-
Chile
Undercurrent
(Fig.
2)
.
From north to south
along its flow path, O
2

decreases along the margin and
reaches levels sufficient to enable subsurface N
-
loss in the
vicinity of 7 to 10
°

S as indicated by the appearance of
NO
2
-

and N deficits (negative N’) and increasing δ
15
NO
3
-
.
CDH
-
23 and CDH
-
26 at 4
°
S are thus located just
upstream
of the low O
2
, N
-
loss
region with cores sites we have
previously studied within a N
-
S gradient of increasing OMZ
and N
-
loss intensity.

Figure 1
.
Locations of cores discussed in this poster.

CDH 23 & 26

3) Dating and Age Models



Unlike cores to the south along the Peru margin
(Fig. 1)
,
forams

were
sufficiently abundant for radiocarbon dating in both CDH
-
23 & 26.
This is likely a consequence that, just outside of the OMZ, bottom
water conditions along the margin remain conducive to carbonate
preservation. Numerous
14
C dates were obtained on both cores to
achieve well resolved age models
(Fig. 3)
. Accumulation rates were
high, ranging from 0.5 to 1.5 m/
kyr
. Also in contrast to previously
studied margin cores, sedimentation was continuous. Whereas
shallower CDH
-
23 reached to the middle of the last deglaciation,
CDH
-
26 spanned the LGM to late Holocene. Benthic
foram

d
18
O
confirms these age assignments. The offset between cores reflects
the warmer overlying water for shallower CDH
-
23.

Figure 3
.
(
A)

Age models for
CDH
-
23 & 26 based on
radiocarbon dating of planktonic
foraminifera. Calendar ages were
derived after subtraction of a
constant
14
C reservoir age effect
(~700
yr
) and calibration using
InterCal

4.0 High and continuous
sedimentation is evident for both
cores.
(
B
)
Benthic
foram

d
18
O vs.
age for each core in comparison to
the
Epica

Antarctic ice core record.
Both the quality of the age model
and the time periods covered by
each core are evident.

A.

B.

4) CDH
-
23 & 26
d
15
N



Modest variations in sediment δ
15
N were
observed over the last 25
kyr
. Where CDH
-
23 &
26 overlap (last 14
kyr
), the records are practically
identical supporting their fidelity in reflecting near
-
surface conditions overlying these two nearby
cores. The most prominent feature is a 1‰
increase between 18 and 14
kyr

followed by a
near
-
steady decrease of 1.5‰ to the late
Holocene. Core top values of 5‰ appear to
represent the modern δ
15
N average, but EUC
source waters actually have a δ
15
NO
3
-

of ~6‰.
The difference is likely due to HNLC conditions at
the site of CDH
-
23 & 26.


Figure
4
. Sediment
d
15
N records for CDH
-
23 & 26.

5
) Comparison to OMZ



d
15
N Records


Within the Peru
-
Chile OMZ, margin sites between
9 and 30
°

S are marked by a large and sharp rise
in δ
15
N. The increase is from 4 to 6‰ and takes
place, at most over 2
kyr
. The CDH 23 & 26
cores, in contrast, have a deglacial δ
15
N increase
that is a fraction of this magnitude and takes twice
as long. The high temporal resolution of these
cores indicates that record quality cannot explain
the difference. We do note that both cores see
subsequent decreases in δ
15
N into Holocene
though there are also clear differences.


Our major conclusion
is that the CDH 23 & 26
reflect changes in the
δ
15
NO
3
-


of the system
and/or local HNLC conditions. Thus most of the
δ
15
N signal found within the OMZ is generated by
changes in OMZ intensity and corresponding N
-
loss. The early rapid rise in OMZ δ
15
N thus
represents a corresponding rapid increase in N
-
loss.

Figure 5
. Sediment
d
15
N records for cores with the OMZ along the Peru
margin (
Fig. 1
).
GeoB

7139 is at 30
°
S and not shown on Fig 1. These data
were previously published by De Pol
-
Holtz (2006,
Paleoceanography
)

Figure
2
.
Biogeochemical maps of the Peru OMZ from
January
and February of
2009. Properties are shown along a constant density surface (
σ
θ

= 26.3 kg m
-
3
)
within the upper portion of the OMZ corresponding to a depth range of 100 to 170
m (Altabet et al., 2012,
Biogeosciences
).
(A)

O
2

concentration (µmol kg
-
1
), station
locations
(
B)

NO
2
-

concentration (µmol kg
-
1
).
(C
) Nitrogen anomaly


N’ (µmol kg
-
1
) calculated as [NO
3
-
] + [NO
2
-
]


16 x [PO
4
-
3
].
(D)

The δ
15
N of NO
3
-
. When
southward intensification of suboxic conditions reaches [O
2
]
<3
µmol kg
-
1
, the
onset of N
-
loss processes is
evident.