Estimates of denitrification in sediments of the Bering Sea shelf

Deep-Sea Research, Vol. 26A, pp. 409 to 4I 5 Pergamon Press Ltd 1979. Printed in Great Britain

Estimates of denitrification in sediments of the Bering Sea shelf ISAO KOIKE* and AKIHIKOHATTORI* (Received 26 June 1978; accepted 1 October 1978)

Abstraet--Denitrification, i.e. anaerobic reduction of nitrate or nitrite to gaseous nitrogen, in the surface sediments of the Bering Sea was estimated using a 15N-tracer method. N 2 production is apparently controlled by the supply of nitrate and nitrite to the sediments. The average rate of N2 production in three locations was 1.2ng-atoms N (g dry weight of sediment)-1 h -1. The rate of nitrate reduction estimated from the vertical distribution of nitrate in the sediments using a one-dimensional diffusion model agreed well with observed rates of N2 production. The annual loss of combined nitrogen by denitrification in the Bering Sea shelf was estimated to be 5 x 101~g. INTRODUCTION COMBINED nitrogen is introduced into the ocean by river runoff and by direct precipitation on the sea surface and is removed by burial in the sediment. Budgets for combined nitrogen in the ocean presented by EMERY, ORR and RITTENBERG (1955), TSUNOGAI (1971), and HOLLAND (1973) all indicate that the input of combined nitrogen exceeds its output. If a steady state is to be maintained, this excess nitrogen must be balanced by biological denitrification. The oxygen-depleted zone of the eastern tropical Pacific has been considered to be a major site for denitrification, and its extent has been assessed by CLINE and RICHARDS (1972), CLINE (1973), and CODISPOTI and RICHARDS (1976). The upper layer of coastal sediments, rich in organic materials and poor in oxygen, is another possible site for denitrification. Oversaturation of dissolved N 2 in the interstitial water of sediments has been observed in basins bordering the southern California continental shelf (BARNS, BERTINE and GOLDBERG, 1975). Using a 15N tracer technique, KOIKE and HATTORI (1978a) presented direct evidence for active denitrification in inshore sediments of the Pacific coast of Japan. The present communication extends this work to the offshore area of the continental shelf of the Bering Sea. The rates of N z production measured during shipboard experiments using 15N-nitrate is consistent with those estimated from the distribution of nitrate in the sediment and overlying water by applying a one-dimensional diffusion model. MATERIALS AND METHODS Sediment samples were collected with a Phleger corer at three stations on the continental shelf of the Bering Sea (Fig. 1) on leg 1 of cruise KH-75-4 of R.V. H a k u h o M a r u (June 21 to July 13, 1975). The hydrographic data have been published by HATTORI (1977). Sediment cores were sectioned to 2-cm thick segments from the top to 10 cm and *Ocean Research Institute, University of Tokyo, Nakano, Tokyo 164,Japan. 409

410

ISAO KOIKEand AKIHIKOHATTORI

....: f ' ~ ' ~'" ..,e.:'-:"'"

"

'W

'70°

,~0"

Fig. 1. Location of stations in the eastern Bering Sea.

into 3-cm thick segments below 10 cm. Interstitial waters of the sediments were extracted using gas-operated squeezers as described by REEBURGH (1967). Within several hours after sampling, N z production from nitrate (denitrification) by bacteria in 1 g (wet weight) of the top 2cm was measured using the 15N method of KOmE and HATTORI (1978a). The rates given in Table I are based on dry weight. Table 1. Rates of Nz production in surface sediments. Location

Date

Temperature

NO3+NO2 in interstitial

N 2 production

waters

(°C)

(ug atoms N 1 -I)

(ng atoms N g-i h-l),

Bering Sea

Sta. 12

9 July 1975

2,5

7.9

1,4

Sta. 14

i0 July 1975

2,5

2.9

0.91

Sta. 19

13 July 1975

2.5

3.9

1.3

8.0

1.6

2.0

8.0

5.9

4.2

3.5

6.2

6.4

3.5

22.2

8.9

Tokyo Bay % (35°34'N, 139°53'E) Sta. 1

19 Feburary 1975

Sta. 2 Mangoku-Ura % (38°25'N, 141o24'E) Sta. 2 Sta. 3

iS January 1976

*g Dry weight of sediments; + I. KO1KEand A. HATTORI,unpublished.

Nitrate plus nitrite in the interstitial waters was estimated by the method of WOOD, ARMSTRONG and RICHARDS (1967) with some modification. After 5 ml of the sample water were passed through a small Cd-Cu column (10cm long and 0.8era in diameter), the column was washed twice with 5-ml volumes of 2 mM EDTA tetrasodium salt solution. The sample and washes were combined and nitrite was determined colorimetrically. Known amounts of KNO3 were dissolved in deionized water, processed in the same way, and used as standards.

Denitrification in Bering Sea sediments

411

RESULTS

Vertical profiles of nitrate (including nitrite) in water column and sediment at Stas 12, 14, and 19 were similar (Fig. 2). There was sharp chemocline at a depth of 30 to 40m; above this depth nitrate concentrations were 0 to 7j.tg at. N 1-1. In the top few centimeters of the sediments nitrate concentrations (3 to 8 ~ag at. N 1-1) were invariably lower than in the overlying water (12 to 29~tg at. N 1- i). The sediments evidently act as sink for nitrate. (m)

100

50 c-

10

2O

(cm)

I

0

I

10 20 NO~-I- NO~2 ( p g atoms N 1-1 )

30

Fig. 2. Vertical profiles of nitrate (including nitrite) in water and sediment at Stas 12, 14, and 19. Averages of nitrate concentrations in the sediment integrated over 2-cm thick layers (0 to 10cm) or 3-cm thick layers (below 10cm) are shown.

Under our assay conditions, N 2 was produced from nitrate in the surface sediments almost linearly with time (Fig. 3), and the capacity for N 2 production was similar, regardless of location. The rate increased with increased nitrate concentrations and was half saturated at ca. 4/.tg at. N 1 - 1 of nitrate (Fig. 4). Concentrations of nitrate in the top sediments were less than 10~tg at. N 1-a. Therefore, in situ N2 production is apparently controlled by nitrate. Estimates of in situ N 2 production in the Bering Sea ranged from 0.9 to 1.4 ng at. N g - 1 h - 1 and averaged 1.2 (Table 1). These values were of the same order of magnitude as those obtained with sediments from Tokyo Bay and Mangoku-Ura on the Pacific coast of northern Japan, although the latter was several times greater (Table 1). DISCUSSION

The vertical distribution of nitrate in the interstitial water of sediments is controlled by a combination of physical processes such as diffusion and sedimentation and biological processes such as microbial reduction or production of nitrate. In coastal sediments rich in organic materials, the rate of anaerobic nitrate reduction far exceeds that of the oxidation of ammonium to nitrite and nitrate (KOIKE and HATTORI, 1978b). The sharp decline in nitrate with depth in the sediments (Fig. 2) indicates that this also holds true for

412

ISAO KOIKE a n d AKIHIKO HATTORI

t-

T¢~4O

40

40

//

z St.14

St. 12

St. 19

o t~ E ~20 c .(2_

2o

"o

o

13.

0 0

i

L

8

16

0 0

i

t

8

16

0

0

I

I

8

16

Incubation time ( h )

Fig. 3. Time courses of N 2 production from tSN-nitrate in the top sediments (0 to 2cm) at Stas 12, 14, and 19 on a dry weight basis. Activities were measured at 2.5°C in the presence of 12lag at. N 1-1 of ~SN-KNO3 (50?/o 15N)"

St. 12

z

1.5

St. 14

o / / / ~ E

1.5

o

o~ 1.0

1.0

r

g •--- 0.5

0.5

"o

:£

'

,

I

5

10

15

0

0

t0

i

i

20

30

NO~ ( pg atoms N 1-1) Fig. 4.

Effects of nitrate concentration o n N 2 production in the top sediments (0 to 2cm) at Stas 12 and 14 on a dry weight basis. Incubation time: 16 h, temperature: 2.5°C.

the Bering Sea shelf sediments. Therefore, we will assume that only bacterial nitrate reduction is quantitatively important. Decreased capacity for bacterial denitrification (a representative of anaerobic nitrate reduction processes) with depth accompanied by decreases in numbers of viable denitrifying bacteria has been demonstrated in inshore sediments (KoIKE and HATTORI,1978a). O u r preliminary survey also showed that the capacity decreases exponentially with depth. The N 2 production is proportional to nitrate concentrati6n (Fig. 4). Thus, the change in nitrate concentration in interstitial water (c) can be represented by the equation: t3c

D t?2c

~t =

c~c

~z 2 - W ~ z

c'Kd°'exp(-Az),

(1)

where D is the diffusion coefficient of nitrate, W is the sedimentation rate, K d ° and A are constants for bacterial nitrate reduction, and z is the depth taken downward. In a steady state, we obtain: ~2C

~C

D ~ z 2 - W ~ z - c" K d ° ' e x p ( - A z )

= 0

(2)

Denitrification in Bering Sea sediments

41 3

The sedimentation rate (W) in the Bering Sea is 0.01 to 0.03cmyear-~ (SHARMA, 1974). MANHEIM (1976) gave a value of 1 × 10 -3 cm 2 s-~ for the diffusion coefficient of nitrate in surface sediments. VANDERBORGHT and BILLEN (1975) used a value of 2 x 1 0 - S c m 2 S-1 for surface sediments of the North Sea coast. We used the latter value. If we assume the balance of the first and second terms of equation (2), using D = 2 × 1 0 - 5 c m 2 s - 1 and W = 0.03 cm year- 1, the decrease in nitrate concentration with depth between 0 and 30cm is less than 1%. Judging from the vertical profiles of nitrate, therefore, biological activity is mainly responsible for the loss of nitrate in the sediments of the Bering Sea shelf, and the second term can safely be neglected. Thus, we obtain :

~2c D ~zSz2 -

= 0.

c. Kd°'exp(-Az)

(3)

Solution of equation (3) for the boundary conditions Z = O~

C = Co (0 ~< c. < Co)

z=z~,c=c.

yields:

1 c = Co/°(2/8) <(C)

~ =~

-2Az-z

exp

B = A \Ed

/

where Io is the zero order modified Bessel function of the first kind (WHITTAKER and WATSON, 1965). Curves fitted to the nitrate data are shown in Fig. 5 together with computer-selected values for the parameters K d ° and A. For the calculation, the nitrate NO~,I-NO~ ( yg atoms N 1-1) 0

5

10

20

i

I

" . ~ 5 t .

12

30 0

tO

,..-.,

"~

A

0.70

10 I

"

Kd" 4.6x10 -5 ~-

5

'

0

10 1

20

3O

I

ol

St. 19 iI

St. 1/-, Kd o

A

1.2x10-4 0.54

Kd*

A

2.1x10-~ 1.3

15

Fig. 5. Vertical distribution of nitrate calculated using equation (3) and values for K d ° and A in the sediments at Stas 12, 14, and 19. Observed values (o) are also shown.

414

ISAO KOIKEand AKIHIKOHATTORI

concentration in the bottom water is used for c~.. Discrepancies between observed and calculated values for nitrate concentrations are significant at only two depths (3 and 5 cm) at Sta. 14. The rates of nitrate reduction at 1-cm depth were calculated using the values obtained for Kd °, A, and c and are represented on the basis of sediment dry weight, assuming that the water content of the sediment is 70~ and the specific gravity is 3 (Table 2). The nitrate reduction rates agreed within a factor of two with the N z production rates in the top sediment (0 to 2cm) (Table 2), suggesting that N 2 is the main product of nitrate reduction. Table2.

Compariso,~'nitratereductionand N2productioni, thesediments~'theBeringSeashe!ll N 2 production

Calculated

at 0 - 2 cm

reduction

(ng a t o m s

N g-i

h-l)

Sta.

12

1.4

0.84

Sta.

14

0.91

0.69

Sta.

19

1.3

2.3

1.2

1.3

Average

nitrate at 1 cm

Nitrate in interstitial water is undoubtedly supplied from overlying waters. Concentrations of nitrate in the bottom waters of the eastern Bering Sea in summer range from 10 to 30j,tg at. N 1-1 (HATTORI, 1977), and winter values (C. P. M c R o f , personal communication, Glacier Cruises 770 and 772) fall in the same range. The average concentration in bottom waters at Stas 12, 14, and 19 was 21 ~tg at. N 1- 1. The difference in the temperature of the bottom water between summer and winter is less than 4°C. If denitrification occurs only in the top 2 cm of the sediments, the annual N 2 production per unit area can be calculated from the average N 2 production rate of 1.2 ng at. N g 1 h (Table 2) to be 1.1 g N m -2. Bacterial denitrification probably occurs over half of the Bering Sea shelf, where the sediment consists of fine sand or silt (SnARMA, 1974). The annual loss of combined nitrogen by denitrification turns out to be 5 x 10 ~ gN. This value is about one twentieth of the annual net supply of nitrate on the Bering Sea shelf estimated as the difference between input and output of nitrate (9 × 1012g N; HAT~ORI and WADA, 1974). If denitrification occurs at the same rate over half of the sea areas shallower than 200m in the world, nitrogen losses in oceanic sediments of 1.5 x 101-~g N year-~ would be expected. This value is nearly equal to denitrification in the tropical eastern North Pacific (1.6 × 1013 g N year- 1 ; CODISPOTI and RICHARDS, 1976). Bacterial denitrification in the sediments is evidently an important factor in the nitrogen budget in the ocean. Acknowledgements We wish to thank the scientists and staff members aboard R.V. Hakuho Maru during cruise KH-75-4 for their co-operation. We are also in debt to Dr M. ENDO for his valuable advice and help in computer analyses and to Dr F. A. RICHARDSfor his critical review of the manuscript. This work was supported by grant (948040) from the Ministry of Education, Culture and Science, Japan.

Denitrification in Bering Sea sediments

415

REFERENCES BARNES R. O., K. K. BERTINE and E. D. GOLDBERG 0975) N 2 :Ar, nitrification and denitrification in southern California borderland basin sediments. Limnology and Oceanography 20, 962 970. CLINE J. D. (1973) Denitrification and isotope fractionation in two contrasting marine environments: the eastern tropical North Pacific Ocean and the Cariaco Trench. Ph.D. Thesis, University of California, Los Angeles. 270 pp. CLINE J. D. and F. A. RICHARDS0972) Low oxygen concentrations and nitrate reduction in the eastern tropical North Pacific Ocean. Limnology and Oceanography, 17, 885-900. CODISI'OTI L. A. and F. A. RICHARDS(1976) An analysis of the horizontal regime of denitrification in the eastern tropical North Pacific. Limnology and Ocea,ography, 21,379 388. EMERY K. O., W. L. OaR and S. C. RITTENBERG 0955) Nutrients in the ocean. In: Essays in natural sciences in honor o/Captain Allan Hancock. University of Southern California Press, Los Angeles, pp. 299-310. HATTORI A. and E WADA (1974) Assimilation and oxidation-reduction of inorganic nitrogen in North Pacific Ocean. In: Ocea,ography of the Beri,g Sea, D. W. HOOD and E. J. KELLEY, editors, Institute of Marine Science, University of Alaska, pp. 149-162 HATTOR1 A. (1977) Preliminary report of the Hakuho Maru Cruise KH-75-4. Ocean Research Institute, University of Tokyo, Tokyo. 87 pp. HOLLAND H. D. 0973) Ocean water, nutrients and atmospheric oxygen. In: Proceedings of the symposium on hydrogeochemistry a,d biogeochemistry, E. INGERSON,editor, The Clarke Co., Vol. l, pp. 68-81. KOIKE I. and A. HATTORI (1978a) Denitrification and ammonia formation in anaerobic coastal sediments. Applied and Environmental Microbiology, 35, 278-282. KOIKE 1. and A. HATTORI (1978b) Simultaneous determination of nitrification and nitrate reduction in coastal sediments by an 15N dilution technique. Applied and Environmental Microbiology, 35, 853-857. MANHEIM [:. T. (1976) Interstitial waters of marine sediments. In: Chemical oceanography, J. P. RILEY and R. CHESTER,editors, Academic Press, Vol. 6, pp. 115-186. REEBURGH W. S. (1967) An improved interstitial water sampler. Limnology and Oceanography, 12, 163-~65. SHARMAG. D. 0974) Contemporary depositional environment of the eastern Bering Sea Part I. Contemporary sedimentary regimes of the eastern Bering Sea. In: Oceanography o/'the Bering Sea, D. W. HOOD and E J. KELLEY,editors, Institute of Marine Science, University of Alaska, pp. 517 540. TSUNOGAI S. 0971) Ammonia in the oceanic atmosphere and the cycle of nitrogen components through the atmosphere. GeochemicalJournal, 5, 57-67. VANDERBORGHTJ.-P. and G. B1LLEN 0975) Vertical distribution of nitrate concentration in interstitial water of marine sediments with nitrification and denitrification. Limnology and Oceanography, 20, 953 961. WHITTAKER E. Z. and G. N. WATSON0965) A course o/modern analysis. Cambridge University Press, pp. 608. WOOD E. D., F. A. ARMSTRONG and F. A. RICHARDS (1967) Determination of nitrate in sea water by cadmium-copper reduction to nitrite. Journal o)"the Marine Biological A ssociatio, o/.the United Kingdom, 47, 23-31.