Mar 11, 1981 - Richard A. Jahnke,' Steven R. Emerson, and James W. Murray ... G. Thompson for assistance in collection .... T. G. Thompson cruise 127. 7. 0.5.
Limnol. Oceanogr., 27(4), 1982, 6 10-623 @ 1982, by the American Society of Limnology
and Oceanography,
Inc.
A model of oxygen reduction, denitrification, matter mineralization in marine sediment? Richard
A. Jahnke,’
School of Oceanography
Steven R. Emerson, WR-IO, University
and organic
and James W. Murray
of Washington,
Seattle 98195
Abstract Porewater nutrient profiles are sensitive indicators of organic matter degradation in sediments. To interpret these data, we have derived a kinetic model of oxygen reduction, nitrification, and denitrification which quantitatively relates the rates of oxygen reduction and denitrification to the depth separating these proccsscs and the carbon oxidation rate. We have compared model profiles with porewater profiles of the sum of NC)-, N&--, and NII,+ (ZN) from both Atlantic and Pacific Ocean stations. Calculated rates of oxygen reduction and denitrification vary proportionally at these stations. Extrapolation along this relationship predicts a maximum in NO3- even where the nitrification zone is so thin that only a decrease from bottom water samples is measured with a l-2-cm sampling interval. The model also predicts that the maximum amount of organic matter that can be oxidized by denitrification is only 30% of that oxidized by oxygen respiration in the Pacific and only 13% in the Atlantic
Within the last 3 years, large amounts of deep ocean porewater nutrient data have been reported (Grundmanis and Murray in press; Emerson et al. 1980; Froelich et al. 1979; Bender et al. 1977). The interpretation of these profiles has been in terms of establishing the stoichiometric relationships between clectron-accepting species and the released metabolites and of determining the order in which the processes of respiration utilize these electron acceptors. Those workers attempted to describe the relative changes in nitrate and alkalinity observed in porewaters on the basis of assumed reactions involving oxygen, nitrate, and MnOz as the terminal electron acceptors and organic matter of Redfield composition. Their studies showed that the oxidants are consumed in the order of decreasing amounts of free energy released during reduction; i.e. oxygen, nitrate, and then manganese, with some
’ Support for this work was provided by NSF IDOE (MANOP) OCE 78-25280-l and NSF grant OCE 78-18746-1. ’ Present address: Scripps Institution of Occanography, Univ. Calif., San Diego, La Jolla 92093.
overlap between the nitrate and manganese reduction zones. Although useful, interpretations such as these do not depend on the rates of the processes and the actual shapes of the profiles. Such information is needed to evaluate the magnitude of elemental sinks within the sediments and to estimate the concentration gradients near the sediment-water interf ace. Our purpose here is to extend the model of Vanderborght and Billen (1975) to deep ocean porewaters by adding an oxygen balance to their equations. The resulting relationships quantitatively relate the rates of oxygen reduction and denitrification to the shapes of the profiles and the amounts of carbon oxidized by these processes. We will first present the model and compare the calculated profiles to those measured at several different areas of the eastern equatorial Pacific and central Atlantic oceans. We then discuss the ramifications and predictions of this model, and methods by which it can be improved and tested. We thank F. Goloway and M. Bender for discussions and criticisms. We also thank the captain and crew of the RV T. G. Thompson for assistance in collection of the Pacific Ocean samples.
610
Respiration The model This is an elaboration of the model of Vanderborght and Billen (1975) in which the oxidation of organic matter by oxygen and nitrate is the dominant reaction affecting the distribution of nitrate. We assume that oxygen respiration is described by Eq. 1 and denitrification by Eq. 2. I38 02 + (CHzO)los(NH3)lsH3P04 + 18 HCO,- + 124 CO, + 16 NO,- + HP042+ 140 H20. (1) 94.4 NW + WLO)mWWJ JW’O, + 13.6 CO, + 92.4 HCO,- + 55.2 N, + 84.8 H,O + HPOd2-. (2) To use Eq. 1 and 2 quantitatively to link the distribution of oxygen and nitrate and the oxidation of organic carbon, we must assume that there is no buildup of reaction intermediates and that the reactions proceed to completion so that our model will describe the distribution of nitrate only. However reaction intermediates of nitrification and denitrification do accumulate in porewaters (Emerson et al. 1980; Suess et al. 1980; Table 1). To account for these intermediates when comparing the model to field data, we define the sum of the measured nitrate, nitrite, and ammonium porewater concentrations as CN. This parameter is used in the discussion instead of nitrate alone. Following Vanderborght and Billen (1975), we partition the sediments into two layers. In the upper region, extending from the sediment-water interface to some depth, Z,, oxygen is the terminal electron acceptor. Below Z,, nitrate is used to oxidize carbon. The equation describing NO,- in the oxygen reduction zone is aN03- = DN)2N03+ k n (3) dt f3Z2 where Dlv is the effective diffusion coefficient of NO,- and k, is a zeroth-order production flmction. The use of a zerothorder term to describe nitrate production during oxygen reduction is only a simple approximation. Among the many factors
611
in sediments
affecting this rate are the concentration of organic carbon and the bacterial population. Berner (1976) modeled the degradation of organic matter in anoxic sediments as a first-order process with respect to organic carbon. For most of the stations investigated here, the oxygen reduction layer is only a few centimeters thick. Since organic carbon profiles for deep ocean sediments are reported for sampling intervals 21 cm, the structure of the organic carbon profile in this layer is not now known. Also, if the rate of bioturbation is large in the upper few centimeters relative to the rate of mineralization, the organic carbon concentration will be relatively constant. In this situation a first-order term will reduce to a constant value. Thus, although a zerothorder production function is an oversimplification, we are forced to retain it until more data become available. For depths greater than Z,t, NOs- is described by
aNO,at
= D a2N03-
- kd NO:,(4) N az2 where k, is the first-order denitrification rate constant. The boundary conditions used to solve these equations are: at Z = 0, NO,- = NOJBW) where NO,-(BW) is bottom water NO,- concentration; at Z = ~0, NO,- = 0; at Z = Z,,, the concentration and first derivative of this expression are continuous. Assuming steady state, the solutions to these equations are (Vanderborght and Billen 1975) for 0 d Z d Zn: NO,- = &-Z2
+ AZ + NO,-(BW)
N
where A=
for Z > Z,l:
(5)
Jahnke et al.
612
Table 1. Nitrate. nitrite. and ammonium norewater distributions for Pacific Ocean stations occupied on cruises TT 127 and TT 145. All units are brnol.kg-l. Dashed lines mark boundary below which sulfate reduction is assumed significant. sta.
7
8
11
12
13
Depth (4
NO,-
NO,-
47.5 47.5 41.9 41.1 38.3 36.9 37.0 34.9 34.0 32.8 32.1 30.9 30.4
0.38 1.79 0.39 0.41 0.43 0.37 0.22 0.22 0.20 0.20 0.17 0.20
1.0 3.5 6.5 9.5 12.5 15.5 18.5 21.5 24.5 30.5 33.5 36.5
43.0 47.2 46.2 43.0 40.5 38.6 38.0 38.7 38.5 37.8 35.8 37.9
1.70 0.44 1.02 0.34 0.34 1.45 0.49 0.19 0.33 0.21 0.27 0.19
0.62 36.2 1.5 6.15 4.5 10.8 3.92 6.4 7.5 2.67 5.0 10.5 1.30 2.9 13.5 0.42 1.5 16.5 0.28 1.3 19.5 0.23 22.5 1.2 ____-----_ _________-_____-_-_-_____ 0.24 0.2 25.5 0.10 0.1 28.5 0.06 0.1 31.5 0.09 0.2 34.5
6.5 6.1 6.7 3.7 7.0 0.7 2.8 ___----5.2 7.5 7.8 -
18.4 0.63 41.2 1.0 3.7 8.26 7.5 3.5 3.5 4.97 2.1 6.5 1.6 1.91 1.1 9.5 1.7 .0.13 12.5 0.76 __________ _____________-____-------------------------
1.
Continued. Depth (4
15.5 18.5 21.5 24.5 27.5 30.5 33.5 36.5
NH,f
T. G. Thompson cruise 127 0.5 0.95 20.5 2.5 1.87 10.0 1.05 5.5 4.5 8.5 0.62 1.1 11.*5 0.16 0.9 _________-_---_---------14.5 0.5 0.13 20.5 1.3 0.09 0.08 26.5 0.12 0.10 29.5 0.55 0.11 38.5 0.19 0.5 2.5 5.5 8.5 11.5 14.5 17.5 20.5 23.5 26.5 29.5 32.5 35.5
Table
0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10
14
13.8 1.5 0.4 4.5 -_---------_____-----_-0.1 7.5 0.1 10.5 0.1 13.5 0.6 16.5 0.1 19.5 0.3 22.5 25.5 0.8 0.1 28.5 0.1 31.5
17
11.1 1.0 0.6 3.5 _____------_ __-_-___----0.4 6.5 0.3 9.5 0.4 12.5 0.5 155 0.8 18.5 0.6 21.5 0.5 24.5 0.8 27.5 T. G. Thompson O-2 44.5 47.9 6-8 49.8 10-12 49.9 14-16 50.2 18-20 50.1 22-24 49.7 26-28 49.5 30-32 49.4 34-36 48.8 38-40 48.7 42-44 48.6 46-48 51.2
NO,-
NH,+
0.04 0.12 0.19 0.07 0.11 0.09 0.08 0.08
:*“o 5:2 8.0 8.6 9.4 10.2 9.5
2.1 5.66 1.0 2.71 _____-__--------------3.2 0.10 4.4 0.10 6.4 0.11 7.2 0.31 7.1 0.11 7.8 0.11 8.2 0.11 8.3 0.11 0.11 8.8 1.13 0.07 0.05 0.05 0.07 0.09 0.10 0.10 0.10 0.14 cruise
10.9 2.7 4.3 7.2 9.6 12.4 13.2 15.0 15.6 15.6
145 -
-
3.2 44.1 N.D. 2.6 . 36.7 N.D. 0.9 4-6 32.0 0.20 1.0 6-8 26.5 N.D. 2.6 8-10 24.0 0.40 1.1 lo-12 20.7 0.40 1.7 14- 16 14.7 0.20 1.3 18-20 8.0 0.30 1.3 22-24 3.9 0.30 2.9 N.D. N.D. 2628 _________________-__---------------------------------3.6 N.D. N.D. 3633 36-39 N.D. N.D. 5.7 42-45 N.D. N.D. 7.3 48-51 N.D. N.D. 9.3 O-2
Respiration
condition at 2, could be changed to a finite value if required but is assumed to be zero here for mathematical simplicity. At steady state, the solution to Eq. 9 is forOsZZ&:
where NO,-,,, (
613
in sediments
=
&N
)
Zn2 + AZ, + N03-(BW).
(8) _
These solutions neglect the possible production of nitrate from the oxidation of ammonium diffusing upward from the sulfate reduction zone. Although an important contributor to nitrate in highly reducing areas, this process is generally negligible at the stations investigated here. For each layer, there is one equation and four unknowns: NOtj-, k,, and k,!, and 2,. Vanderhorght and Billen were able to reduce the number of variables at this point by using measurements of nitrate in the porewater and of the dark uptake rate of [‘4C]bicarbonate to estimate k,. Since we know of no measurement of the latter in deep ocean sediments, we will include oxygen considerations to reduce the number of variables. The rate at which oxygen is consumed is proportional to the rate at which NO,is produced (Eq. 1). Thus, the distribution of oxygen in the 0, reduction zone can be described by
(9) where Do is the diffusion coefficient of 0, in porewater and y is the 02:N ratio in Eq. 1, assumed here to be 138:16 or 8.625. The boundary conditions for oxygen are: at 2 = 0, 0, = 02(BW) where O,(BW) is the bottom water concentration; at 2 = Zll, O2 = 0. The latter condition states that all of the oxygen is consumed before the onset of denitrification. Although this is an oversimplification, it is well established both in the field and laboratory that oxygen is significantly depleted before the onset of denitrification (Richards 1965; Ozretich 1976). The error introduced into the model by this assumption is, thus, small. The boundary
yb2k2
+
200
1
02&W
z
2Dozn
[
+ 02(BW).
00) An additional constraint in our model is that the flux of oxygen across the sediment-water interface must equal the amount consumed.
-D ao, Znk, (‘az Iz=() = I () Evaluating this expression rect relationship between 2% [
02OW ykt
dZ.
(11)
results in a dik, and Z,:
1
t =
z ?L ’
(12)
At this point, the number of variables has been reduced to three: NOs-, k,, or Z,I, and kd. To help evaluate these variables we consider the relationship between 0, and NOs- consumption and organic carbon oxidation. The amount of organic carbon oxidized by oxygen (Co) at steady state is equal to the amount of oxygen consumed in the upper zone (0 < Z < Z,) multiplied by the carbon to oxygen ratio in Eq. 1: (13) The quantity of organic matter oxidized by denitrification, CN, is qua1 to the flux of nitrate through the Z = Z, horizon multiplied by the carbon to nitrate ratio in Ey. 2:
The total amount of carbon consumed in the sediments by these processes is the sum of Eq. 13 and 14 and is referred to ilS CT. At many stations in hemipelagic areas
614
Jahnke et al.
resent progressively faster rates of denitrification. As the rate of denitrification increases the maximum concentration &2 D 6702-yk n: St OS22 decreases, until eventually no maximum pJLcDNm+ k” exists (f and g). Similarly, the depth of the St St2 concentration maximum decreases with increasing denitrification rates until the DEN rRlC KArlON ZONL I TO tmaximum disappears altogether. The ki depth at which oxygen goes to zero, Z,, ~3 25 is the same for all of these profiles. A decrease of N03- in the oxygen reduction 35 zone results solely from diffusion downward and not from consumption. This is i 40 in contrast to the model of Bender et al. Fig. 1. Example model profiles calculated for (1977) in which the depth of the nitrate O,(BW) = 110 r.Lmol*kg-‘, NO,-(BW) = 40 prnol. kg-‘, DN = 3.5 x lo-” cm2. s -‘, IJo = 4.2 x 10P6 maximum is considered to mark the excm2.8 ‘, and k, = 1 X lo-l4 prnol. cln-“. s-‘. Values haustion of oxygen. of k,, used for each profile are: a-lO-‘2a s-l; I+ The decrease in the depth and magniI() 10. s-l; c-lo-~.s-‘; d-10-8. s 1; -10 7-s-l; ftude of the concentration maximum with lo-“*s ‘; g-10-4. s-“. increasing rates of denitrification (Fig. 1) is quantitatively evaluated in Fig. 2. The variation in the ratio of the depth of the of the deep sea the processes of nitrifimaximum, Z,,,, to the depth at which cation and denitrification are sufficiently oxygen becomes zero, Z,, is presented as rapid so that the nitrate profile is coma function of log kJk, in Fig. 2A. The pressed into a thin zone in the surface abscissa, log kJk,L is a relative scale with sediments and cannot be adequately rerelatively slow rates of denitrification to solved with a sampling interval of 2-3 the left and increasing to the right. Where cm. It is helpful at these stations to dislog kJk,, is ~3 cm3. molll, the Zmax:Z,L racuss the processes in terms of the total tio is nearly 1.0, indicating that the NO,standing stock of nitrate, (NO,-),, rather maximum occurs very near the depth at than attempt to evaluate the curvature in which oxygen becomes exhausted. It is the profiles. (N03-)r is the integrated only under these conditions that the stoiamount of NO,- in the sediment column, chiometric model proposed by Bender et al. (1977) is correct. However, for the sta(NO,-), = I* (NO:,-) dz. (15) tions considered here, the rate of denitri0 fication is slow enough that these condiThe solution of this expression is . tions are satisfied and the NO,- maximum does occur near the depth of O2 exhaus(NO,-),,. = w + F tion as assumed by Bender. At values of log k,Jk,, > 4 cm”. mall’, Zlnax:Z,, is signifi+ NO:,-(BW) Z,L cantly
