of 15
Carbonate
dissolution
fluxes in deep-sea
sediments
as determined
from in
situ porewater
profiles in a transect
across the saturation
horizon
Jaclyn E.P. Cetiner
a,*
, William
M. Berelson
a
, Nick E. Rollins
a
, Xuewu Liu
b
, Frank J. Pavia
c
,
Anna R. Waldeck
d
, Sijia Dong
c
, Kalla Fleger
b
, Holly A. Barnhart
c
, Matthew
Quinan
a
,
Rucha P. Wani
a
, Patrick A. Rafter
e
, Andrew
D. Jacobson
d
, Robert H. Byrne
b
, Jess F. Adkins
c
a
University
of Southern
California,
Los Angeles,
CA 90089,
USA
b
University
of South
Florida,
St. Petersburg,
FL 33701,
USA
c
California
Institute
of Technology,
Pasadena,
CA 91125,
USA
d
Northwestern
University,
Evanston,
IL 60208,
USA
e
University
of California,
Irvine,
Irvine,
CA 92697,
USA
ARTICLE
INFO
Associate
editor
: Christof
Meile
Keywords:
Carbonate
dissolution
Porewater
geochemistry
Benthic fluxes
Saturation
state
Carbonate
burial
ABSTRACT
Despite
their importance
for long-term
climate
regulation,
the rates and mechanisms
of seafloor
carbonate
dissolution
are poorly understood,
especially
with respect to calcite saturation
and the role of sedimentary
metabolic
CO
2
production.
Here, we present results from an in situ porewater
sampler
deployed
at the Cocos
Ridge in the eastern equatorial
Pacific, where we examine
seafloor
carbonate
dissolution
in locations
with
bottom water
Ω
calcite
ranging
from 1.0 to 0.84 (1600
3200 m). With cm-scale
resolution
from the sediment
-
waterinterface
to35cm,wepresentporewater
profilesoftotalalkalinity,
pH,dissolved
inorganic
carbon(DIC),
δ
13
C of DIC,
Ω
calcite
, [Mn], [Ca], and [Sr], as well as solid phase porosity,
% CaCO
3
, and % organic C. These
profiles
provide
evidence
that deep-sea
sedimentary
carbonate
dissolution
occurs via sediment-side
control,
wherein
dissolution
isdominated
bysedimentary
processes
ratherthanstrictlybottomwatersaturation
state.We
estimate
dissolution
fluxesusingthreeindependent
approaches:
alkalinity
fluxes,
δ
13
CofDICcombined
withDIC
fluxes, and [Ca] fluxes. We report seafloor
dissolution
fluxes with uncertainties
<
38 %: 40
±
15, 98
±
20, 100
±
32, and 89
±
27
μ
mol CaCO
3
/m
2
/day at sites 3200, 2900, 2700, and 1600 m deep, respectively.
The
magnitude
of dissolution
fluxes is a function
of bottom water saturation
state (
Ω
calcite
), bottom water dissolved
oxygen,
and sedimentary
CaCO
3
content,
but not correlated
with any of these parameters
independently.
We
observe
dissolution
occurring
at all stations,
including
where bottom water is saturated
with respect to calcite,
and present evidence
that this occurs through
respiration-driven
dissolution
within the sediment.
At all sites,
porewater
Ω
calcite
decreases
below bottom water values before increasing
toward saturation
deeper in the
sediment.
Usingthe
δ
13
C ofDIC,wepartition
theDICfluxesacross thesediment
water interface
andfind21
48
% of DIC is sourced
from CaCO
3
dissolution,
with the remainder
sourced
from organic matter respiration.
We
present a sedimentary
mass balance,
assembled
with dissolution
rates and mass accumulation
rates obtained
through
Δ
14
C of foraminiferal
calcite, and calculate
CaCO
3
burial efficiencies
between
2 and 67 %, inversely
correlating
withwaterdepth.Ourresultsalsoprovideevidence
thatnetchemical
erosionof5,000
––
10,000year
old carbonate
is occurring
at the deepest site. Aerobic
organic C respiration
coupled
with sedimentary
CaCO
3
dissolution,
asdocumented
here,willprovidemorealkalinity
tobottomwatersthanfromundersaturation-driven
dissolution
alone. This process can neutralize
anthropogenic
CO
2
at the seafloor
in a larger range of saturation
states than previously
estimated.
1. Introduction
The ocean has absorbed
roughly
one third of fossil fuel emissions
since the start of the Industrial
Revolution
(Doney et al., 2009; Khati-
walaetal.,2013;Gruberetal.,2019).ThisinfluxofCO
2
hasledtoocean
acidification
with an increase
in surface
seawater
hydrogen
ion
* Corresponding
author.
E-mail
address:
jepittma@usc.edu
(J.E.P. Cetiner).
Contents
lists available
at ScienceDirect
Geochimica
et Cosmochimica
Acta
journal
homepag
e:
www.else
vier.com/loc
ate/gca
https://doi.org/10.1016/j.gca.2024.11.027
Received
6 January
2024; Accepted
26 November
2024
Geochimica
et
Cosmochimica
Acta
390
(2025)
145–159
Available
online
30
November
2024
0016-7037/©
2024
The
Authors.
Published
by
Elsevier
Ltd.
This
is
an
open
access
article
under
the
CC
BY
license
(
http://creativecommons.org/licenses/by/4.0/
).
concentration
(Doney et al., 2020). Based on modeling
studies, ocean
buffering
will neutralize
this excess acidity from anthropogenic
CO
2
on
timescales
of 10,000 years (Archer et al., 2009) via dissolution
of sea-
floor carbonate
sediments.
There is already evidence
of anthropogenic
CO
2
being buffered
by deep-sea
carbonate
dissolution
in some areas of
the North Atlantic
(Sulpis et al., 2018), but it will take on the order of
1000
2000 years, i.e., via global ocean ventilation
(Broecker
et al.,
1984) for anthropogenic
CO
2
to penetrate
throughout
the deep ocean.
Carbonate
(CaCO
3
) dissolution,
particularly
in the deep ocean, is a
natural
mechanism
by which the planet neutralizes
CO
2
. When
anthropogenic
CO
2
arrivesat the seafloor,
eitheras dissolved
CO
2
or as
particulate
organic carbon (POC) following
photosynthesis,
knowledge
of the rate of CaCO
3
dissolution
and the parameters
controlling
disso-
lution will improve
understanding
of how the ocean is responding
to
anthropogenic
climate
change.
The amount
and location
of POC
oxidation,
whether
by oxygen or other oxidants,
is another
important
consideration
of the fate and neutralization
of anthropogenic
CO
2
.
Respiration-driven
dissolution
can enhance
sedimentary
dissolution
through
theproduction
ofacidviaaerobicrespiration
oforganiccarbon.
The subsequent
dissolution
of CaCO
3
neutralizes
CO
2
via the following
coupled
reactions
(Emerson
and Bender, 1981):
C
org
+
O
2
CO
2
; CO
2
+
H
2
O
+
CaCO
3
2HCO
3
-
+
Ca
2
+
(1)
Emerson
and Bender (1981) concluded
that CaCO
3
dissolution
is at
least partially
controlled
by the relative
rain ratios of organic
C and
CaCO
3
. Martin and Sayles (1996) found that significant
dissolution
could be occurring
in sediments
above the saturation
horizon
due to
metabolic
CO
2
production,
basedonorganiccarbonoxidation
ratesand
dissolution
rates estimated
from in situ squeezed
porewater
measure-
mentsofcalcium[Ca],alkalinity,
anddissolved
inorganic
carbon(DIC).
DIC is not only produced
by aerobic respiration,
but also by sub-oxic
diagenetic
reactions
occurring
within the sediment
(Froelich
et al.,
1979).Thedepthofoxygenpenetration
maysignificantly
affecttherate
at which respiration-driven
dissolution
can occur, regardless
of the
bottom water saturation
state (Martin and Sayles, 2006). Additionally,
even if the rain rates of POC are higher than those of CaCO
3
, CaCO
3
dissolution
maynotoccuriftheorganiccarbonis oxidized
very nearto
the sediment
water interface
(SWI), thereby
allowing
bottom water
CO
3
2
to neutralize
metabolic
CO
2
(Sayles et al., 2001). Whether
respiration-driven
dissolution
occurs primarily
at the sediment
water
interface
or within the sediment
column is difficult
to assess without
accurate,
high-resolution
porewater
profiles.
An ongoing
debate regarding
seafloor
dissolution
is whether
disso-
lutionisunderwater-side
orsediment-side
control,i.e.,ifcontrolofthe
dissolution
rate depends
primarily
on the diffusive-boundary
layer be-
tweensediment
andoverlying
water,ortheporewater
chemistry.
Some
models and sediment
bed reactors
indicate
that seafloor
CaCO
3
disso-
lution is entirely
dependent
on boundary
layer conditions
(Boudreau,
2013;Sulpisetal.,2017).Onemajorimplication
ofwater-side
controlis
the prediction
that there is no change in porewater
pH or calcite satu-
ration state (
Ω
calcite
) below the upper few mm of the sediment
column
(Boudreau
etal.,2020).Because
porewater
hasbeendifficult
tosample
and measure
accurately
in the past, no artifact-free,
high-resolution
Ω
calcite
porewater
profiles have previously
been generated
to test this
implication.
Downcore
profiles of pH and other carbonate
parameters,
below the top few mm, would provide
key evidence
of water-side
or
sediment-side
control.
Quantifying
ratesofCaCO
3
dissolution
hasbeenalongstanding
goal
in oceanography,
and several methodologies
have been employed
to
constrain
the marine carbonate
budget, such as benthic landers (e.g.,
Berelson
etal.,1990a;SaylesandDickinson,
1991;Jahnkeetal.,1997),
in situ microelectrodes
(e.g., Archer et al., 1989; Reimers
et al., 1992;
Hales andEmerson,
1997),massbalances
(e.g., Balchet al.,2007),and
modeling
efforts(e.g., Broecker
andBroecker,
1974;Archer and Maier-
Reimer,
1994). In addition
to dissolution
rate, the location
in the
sediment
column where dissolution
occurs is also of interest.
While the
parameters
driving dissolution,
such as saturation
state, bottom water
oxygen concentration,
CaCO
3
mineralogy,
and organic carbon supply
have been the topic of many studies,
uncertainty
remains
about the
relative
importance
of these factors and their control on sedimentary
dissolution.
Recently,
dissolution
kinetics
have been well studied in bench-top
experiments
and the water column to help refine parameterization
of
the dissolution
rate law relating
CaCO
3
dissolution
rate to saturation
state (e.g., Subhas et al., 2015; Dong et al., 2018; Dong et al., 2019;
Naviaux
et al., 2019a; Adkins et al., 2021). An outstanding
question
is
whether
thesekineticratelawsapplytodissolution
withinthesediment
(Boudreau
et al., 2020). To independently
constrain
this question,
bulk
sediment
dissolution
fluxes must be defined,
and here we summarize
how this has been done previously.
Much work has been done to determine
seafloor
CaCO
3
dissolution
fluxes by direct measurements
using benthic flux chamber
incubations.
Near our study site, in situ seafloor
dissolution
fluxes were determined
usingchambers
onlandersinatransect
acrosstheequatorinthecentral
Pacific (Berelson
et al., 1990a, 1990b, 1994). A study by Jahnke et al.
(1997)offthecoastofCalifornia
foundthatmodeled
ratesofcarbonate
dissolution
underestimated
ratesmeasured
inchambers,
indicating
that
metabolic
CO
2
mustsignificantly
contribute
totheobserved
dissolution
rates. The ROLAI
2
D lander captured
fluxes of alkalinity
and DIC in the
NorthAtlantic,
andtheirratiowasusedasaproxyforCaCO
3
dissolution
and respiration
(Sayles and Dickinson,
1991).
Microelectrodes
can provide
high resolution
profiles
that can be
modeled
forestimates
ofCaCO
3
dissolution
withinsediments;
themajor
benefitsofmicroelectrodes
includemm-scale
depthprecision
andinsitu
measurement
capability.
Oxygen
and pH microelectrodes,
in conjunc-
tion with numerical
models,
have been used to calculate
dissolution
rates in the top few cm of sediment
(Archer et al., 1989; Reimers
et al.,
1992) and to measure
oxygen flux into sediment,
a proxy for aerobic
respiration
that could drive dissolution
(Hales et al., 1994). However,
there is uncertainty
in the interpretation
of microelectrode
pH data
given the multiple
reactions
that can affect pH. The use of in situ pCO
2
microelectrodes
also provides
evidence
that a fraction
of sedimentary
dissolution
occurs through
respiration-driven
processes
(Hales and
Emerson,
1997).
The flux of dissolved
calcium
may be considered
a more definitive
proxy for CaCO
3
dissolution
(Jahnke et al., 1997) given the many
sources
and sinks of alkalinity
and DIC, but fewer major reactions
involving
Ca in the shallow
sediment
column.
However,
the measure-
mentprecision
neededtoobserveslightchanges
in[Ca]canbedifficult
to achieve.
Measurements
of excess Ca relative to Na in porewater
has
also been considered
a signal of sedimentary
dissolution
in coastal
sediments
(Greenetal.,1998).Catoalkalinity
ratioshavebeenusedasa
metric of CaCO
3
production
and dissolution
and other cation exchange
processes,
withanalyses
inporewater
(Steineretal.,2022),benthicflux
chambers
(Jahnke and Jahnke,
2004), and the water column (Steiner
et al., 2021).
Insitufiltration
ofporewater
isnecessary
toavoidsampling
artifacts
associated
with extreme
pressure
and temperature
changes
occurring
whencorestravelfromthedeepseatothesurface(Bischoff
etal.,1970;
Bischoff
andSayles,1972;Saylesetal.,1973;Saylesetal.,1976;Murray
etal.,1980;Sauvage
etal.,2014;Cetineretal.,2023).Here,wepresent
the first high-resolution
porewater
profiles (0
35 cm depth range) ob-
tainedfromporewater
extracted
insituatsitesintheCocosRidgeinthe
eastern equatorial
Pacific. We report pH, total alkalinity,
DIC,
δ
13
C of
DIC,and [Ca] measurements
on in situfiltered porewater
and interpret
these profiles via Fick
s first law to estimate
seafloor
carbonate
disso-
lution fluxes using three approaches:
alkalinity
fluxes,
δ
13
C and DIC
fluxes,and[Ca]fluxes.WealsopresentabudgetforCaCO
3
rain,burial,
and dissolution
on the Cocos Ridge seafloor.
J.E.P.
Cetiner
et al.
Geochimica
et
Cosmochimica
Acta
390
(2025)
145–159
146
2. Study
area
and methods
2.1.
Field
site
In December
2021, we conducted
a cruise on the R/V
Sally
Ride
on
andadjacent
totheCocosRidgeintheeasternequatorial
Pacific(Fig.1).
This region contains
a wide range of surface sediment
carbonate
con-
tents:
<
1
93 wt% CaCO
3
(Moore et al., 1973; Cetiner et al., 2023).
Previous
studies have found sedimentation
rates of 1
3 cm/kyr in this
region (Lea et al., 2000; Liao and Lyle, 2014). In a depth transect
of
1600
3200m,weencounter
changes
incalcitesaturation
stateranging
from 1.0 to 0.84 (Table 1).
Waterdepthandbottomwaterproperties
aresummarized
inTable1.
Station1waslocatedfurthestfromtheridgeanddeepest;
Stations
5and
2wereasimilardistance
fromtheridge;Station3wasontheridgeand
shallowest
(Table1).Stations
willbepresented
inthispaperinorderof
deepest to shallowest:
1, 5, 2, and 3. Weight % CaCO
3
correlates
with
depth and distance
to ridge; stations that were shallower
and closer to
theridgehadthehighestsedimentary
carbonate
content(Table1).This
transect
can be thought
of as a section down the modern
sedimentary
lysocline.
At each station, two CTD casts measured
dissolved
oxygen (Fig. 2a)
and Niskin samples
were collected
for analyses
of DIC,
δ
13
C of DIC,
alkalinity,
and pH. These parameters
were all measured
to over-
constrain
the water column
carbonate
chemistry
system and deter-
mine the saturation
state of calcite (
Ω
calcite
) in bottom waters (Fig. 2b)
usingCO2SYS
(v3.1.1).
Thedepthrangeofthefourstationsallowsusto
separately
analyze the effects of bottom water saturation
state and ox-
ygen content on carbonate
dissolution
fluxes.
2.2.
Porewater
sampling
Porewater
samples
were collected
using SIPR, a device we built
capable
of filtering
porewater
in situ from deep-sea
sediments;
its
design, field validation,
and comparison
to traditional
methods
are
described
indetailinCetineretal.(2023).Briefly,twotypesoffiltering
mechanisms,
termed
needles
and
blades,
areattached
toa standard
MC-800
multicorer
and deployed
off a ship in the traditional
manner
except that, once landed, the multicorer
remains
stationary
on the sea-
floor for 6
10
h
. The needles
and blades have
windows
that act as
sampling
ports, holding a Rhizon filter (needles)
or covered
with filter
paper (blades).
The needles
and blades penetrate
the sediment
with
sampling
ports distributed
between
0
35 cm, with high resolution
(0.5
cm spacing)
near the sediment
water interface.
Upon computer
actua-
tionviaburnwires,spring-loaded
syringes
drawporewater
through
the
filters and samples
are stored in 30 mL glass quartz coils designed
to
Fig. 1.
Map of station locations
on the Cocos Ridge. Bold contours
denote every 1000 m; thin contours
denote every 200 m.
J.E.P.
Cetiner
et al.
Geochimica
et
Cosmochimica
Acta
390
(2025)
145–159
147