APPLIED AND ENVIRONMENTAL MICROBIOLOGY, June 1986, p. 1174-1179 Vol. 51, No. 60099-2240/86/061174-06$02.00/0Copyright C 1986, American Society for Microbiology

Temporal Variation of Denitrification Activity in Plant-Covered,Littoral Sediment from Lake Hampen, Denmark

PETER BONDO CHRISTENSEN AND JAN S0RENSEN*Department of Genetics and Ecology, University of Aarhus, Ny Munkegade, DK-8000 Aarhus C, Denmark

Received 19 August 1985/Accepted 24 January 1986

Diel and seasonal variations in denitrification were determined in a littoral lake sediment colonized by theperennial macrophyte Littorella uniflora (L.) Aschers. In the winter, the activity was low (5 ,umol ofN m 2 h-1)and was restricted to the uppermost debris layer at a depth of 0 to 1 cm. By midsummer, the activity increasedto 50 ,umol of N m2 h-1 and was found throughout the root zone to a depth of 10 cm. The root zone accountedfor as much as 50 to 70% of the annual denitrification. A significant release of organic substrates from the rootsseemed to determine the high activities of root zone denitrification in the summer. The denitrification in thesurface layer and in the root zone formed two distinct activity zones in the summer, when the root zone alsocontained nitrification activity, as indicated from the accumulations of N03 . Light conditions inhibiteddenitrification in both the surface layer and the upper part of the root zone, suggesting that a release of 02 bybenthic algae and from the roots of L. uniflora controlled a diel variation of denitrification. In midsummer, therate of denitrification in both the surface layer and the upper part of the root zone was limited by N03 . In thegrowth season, there was evidence for a significant population of denitrifiers closely associated with the rootsurface.

The earliest information on denitrification in freshwaterlakes was obtained from mass balances (2) and from deter-minations of N03- uptake in the sediments (1, 3). Subse-quent investigators used 15N isotope and in situ enclosures tomeasure such uptake rates (22), and the enclosures have alsobeen used in combination with the acetylene inhibitiontechnique to give a direct measurement of denitrification bythe formation of gaseous nitrogen (9, 10). It would seem thatonly those assays, which are based on the direct determina-tion of N20 and N2 production, should be used for themeasurement of denitrification in natural sediments. Forinstance, the assay of uptake of NO3- may also incorporatethe route of assimilation by benthic microalgae and maytherefore lead to an overestimate of the denitrification.Furthermore, the early assays of NO3- uptake included anaddition of NO3- to the water phase, which could haveaffected the rate of N03- reduction (3, 22).

In littoral sediments, nitrogen cycling may be stronglyinfluenced by the root metabolism of aquatic macrophytes(8, 13, 14, 19). Root activities such as the uptake of NO3-,release of 02(6, 13, 17), and excretion of organic compounds(7, 12, 20) may all affect the denitrification in sediments, andboth the spatial heterogeneity and the complicated controlpatterns are serious restraints on the quantification ofrhizosphere denitrification. None of the earlier reports havedescribed the seasonal or diel patterns of bacterialdenitrification in plant-embedded lake sediments.

In the present study, we used the acetylene inhibitiontechnique to measure a full annual cycle of denitrification ina littoral sediment which was densely covered with theperennial isoetid Littorella uniflora (L.) Aschers. Light anddark incubations were used to assess diel patterns. Theregulating factors for denitrification were considered to bethe changing temperature and light conditions as well as theavailability of 02, NO3-, and organic substrates in thesediment.

* Corresponding author.

MATERIALS AND METHODS

Sediment and macrophyte community. Lake Hampen is a

small (76 ha), oligotrophic lake in mid-Jutland, Denmark. Thelittoral vegetation is dominated by L. uniflora (L.) Aschers.,which is about 5 cm tall and forms very dense and almost purestands in the sandy, oxidized sediment from the waterline toa depth of about 1.5 m. Benthic microalgae, especiallydiatoms, are abundant in the littoral zone. In the sediment, a

high organic content is observed within the upper 3 cm (1 to3% as measured by ignition loss) and is distinguished by itsdarker color from the underlying sand, in which the organiccontent is only between 0.2 and 0.5%.A sampling site in the littoral zone was chosen in which

the plant community was well developed with a density ofabout 9,000 plants m2. Core samples were taken at intervalsof 1 to 3 months from August 1983 to December 1984. Oneach occasion, a series of sediment cores (15 by 4.5 cm) weretaken by hand in Plexiglas tubes. On the average, each of thecores contained about 15 rooted plants; all were intact, withtheir roots in the sediment and the above-ground tissues inan aqueous phase of about 200 ml. The plants had a rootzone to a depth of about 10 cm in the cores, but the highestdensity of roots was observed at a depth of 1 to 3 cm. Theassay of denitrification was initiated in the laboratory within2 h of collection of the cores.

Assay of in situ N03- concentration. Separate cores andwater samples were taken for the determination of in situconcentrations of NO3- in both the sediment and the lakewater. The cores were cut into 1-cm segments and trans-ferred to beakers containing 15 ml of 1 N KCl solution. Afterbeing vigorously shaken, the slurries were centrifuged (3,000x g, 10 min), and the supernatants were analyzed for NO3-(5) in an autoanalyzer (Chemlab Instruments Ltd., Essex,England). All concentrations were micromolar and werecorrected for sediment porosity.

Assay of denitrification activity. A slight modification of theacetylene inhibition technique described by S0rensen (21)was adopted for the assay of sediment denitrification. Inshort, the activity is determined as the rate of N20 accumu-

1174

DENITRIFICATION IN LAKE SEDIMENT 1175

lation in cores which are amended with dissolved C2H2 toblock the bacterial reduction of N20 to N2. At depthintervals of 1 cm, an aliquot of 600 ,ud of C2H2-saturated,distilled water was distributed homogeneously into the sed-iment. A 20-ml portion of the overlying aqueous phase wasreplaced with C2H2-saturated water, and a stopper wasmounted on top. The final concentration of C2H2 was ap-proximately 10% of saturation in both the pore water and theoverlying aqueous phase.A total of six cores were amended with C2H2 by this

procedure and incubated from 2 to 6 h at the in situtemperature; three of the cores were left in darkness, and theother three were incubated under the light of a 400 Wmercury lamp (Osram HQI-T). Irradiance was 200microeinsteins m-2 s-' at the sediment surface; this level issufficient to saturate the photosynthetic activity of L.uniflora (16). The temperature variation was 2 to 20°C duringthe year.

After incubation of a core, a subsample of 25 ml was takenfrom the aqueous phase and stored in a glass vial containing200 ,ul of saturated HgCl2 solution. The vial was completelyfilled with the sample and stoppered quickly to avoid a lossof N20. The N20 could later be extracted from the sampleby a simple headspace technique in which 5 ml of the waterwas substituted with an equivalent amount of N2. Aftervigorous shaking of the vial, the N2O content could beanalyzed by gas chromatography, as described below.

After the water samples had been removed, the cores werestoppered again and frozen at -20°C. The following day, thecores were cut into 1-cm segments and quickly transferred to60-ml beakers containing 15 ml of 1 N KCl solution. Thebeakers were immediately stoppered and shaken for 10 minuntil the sediment was thoroughly thawed and N20 was inequilibrium between the aqueous and gaseous phases. Gassamples (3 ml) were taken from the beakers with a syringeand stored in preevacuated glass vials (Venoject; TerumoCorp., Tokyo, Japan). Finally, the slurries were centrifuged(3,000 x g, 10 min), and the supernatants were analyzed forNO3-.

All gas samples were analyzed for N20 on a Packard 427gas chromatograph which was equipped with a63Ni electroncapture detector held at 320°C. A Porapak Q column (2 m by3.2 mm; 80/100 mesh; Waters Associates) was used toseparate the components at 60°C, with pure N2 as the carriergas (flow rate, 15 ml min-1).For the headspace analyses, a correction was included for

the dissolved N2O. Bunsen solubility coefficients of 0.6(water samples extracted at 22°C) and 0.7 (sediment samplesextracted at 16°C) were used (23). Denitrification activitieswere estimated as the linear accumulation of N20 during thetime of incubation (correlation coefficient > 0.85) and ex-pressed in nanomoles of N centimeter-3 hour-' for eachsegment of the sediment. Because of the short incubationtime, the amount of N20 accumulating in the aqueous phaseof the cores was assumed to originate from denitrification inthe surface layer of the sediment and was therefore assignedto the 0- to 1-cm segment. The overall activity in wholesediment cores was expressed in micromoles of N meter-2hour-'. During the incubation period, there were no mea-surable changes in the NO3 concentrations at any depths ofthe sediment.

Assay of denitrification in N03-amended cores. To eluci-date the control of NO3- availability on in situ denitrifica-tion, a separate set of cores was incubated with C2H2-saturated water containing 10 mM KNO3. The final NO3-concentration in the pore water of these cores was between

600 and 900 ,uM, which was well above the in situ concen-trations. All incubations and analyses of the NO3--amendedcores were performed as described for the assay ofunamended cores.

RESULTS

Seasonal variation of N03- concentrations. Except for aperiod in early summer (April through June), when thehighest NO3 concentration was found at a depth of 1 to 2cm, maximum concentration of NO3 was always within theuppermost 1 cm (Fig. 1). The concentration decreasedrapidly with depth, although it was always above 10 ,uM inthe root zone, even at a depth of 10 cm. Concentrations inthe overlying water column were also lower. The concentra-tions were lowest in the summer, but increased again duringthe fall and reached a maximum of 200 to 300 ,uM in thesurface layers during the early spring.A sudden change of the balance between production

(nitrification) and consumption (assimilation and denitrifica-tion) of the N03 was apparent from its rapid decrease inabundance in the spring (Fig. 1). The assimilatory demandfor inorganic nitrogen was expected to be high inmacrophytes and algae at this time of year, which marks theonset of the growth season.

Seasonal variation of in situ denitrification. Under bothlight (Fig. 2A) and dark (Fig. 2B) conditions, the denitrifica-tion showed a marked seasonal variation. During the coldermonths, from October to April, only the dark-incubatedcores showed detectable denitrification rates (0 to 0.1 nmolof N cm3 h-1), and the denitrification occurred only in thesurface zone of the sediment. No activity was observed inthe light-incubated cores during the winter. As the growthseason began, however, there was an increase in activity andan indication of the presence of two separate denitrificationzones during the summer: one in the surface layer at a depthof 0 to 1 cm and one in the underlying root zone. Thesurface-associated activity was again highest in dark-incubated cores (0.1 to 1.0 nmol of Ncm3 h-1) but was nowalso significant in the light-incubated cores (0.1 to 0.5 nmolof N cm3 h-1). Illumination thus resulted in a completeinhibition of the surface activity in the winter, while in thesummer the activity was only inhibited by ca. 50%. The rootzone denitrification, which was not detectable in the winter,showed maximum activity at a depth of about 3 cm in thesummer (0.5 nmol of N cm3 h'1). In this zone, however,core illumination had only a small effect (see below).By vertical integration of the data presented in Fig. 2,

areal estimates of denitrification were obtained (Fig. 3). Thedistribution of activity between the surface zone (depth, 0 to1 cm) and the upper and lower parts of the root zone (depth,1 to 3 and 3 to 10 cm, respectively) is indicated. The datapresented in this way demonstrated (i) the high overallactivity in the summer and (ii) an apparent inhibition ofdenitrification in the light for all seasons. During the summer(August), maximum activities of 40 and 50,umol of Nm-2h-1 were recorded for light and dark incubations, respec-tively. The activities were markedly lower in winter, about 0and 5,umol of N m-2 h-1, for the same conditions. Thedistribution of activity with depth also changed during theyear. In the summer, for instance, the fraction of denitrifica-tion occurring in the root zone represented as much as 40 to80% of the total recorded in the light (Fig. 3A) comparedwith 40 to 60% of the total recorded in the dark (Fig. 3B). Forboth light and dark conditions, only one-third of the overallactivity was found in the upper part, at a depth of 1 to 3 cm.

VOL. 51, 1986

1176 CHRISTENSEN AND S0RENSEN

E0

m

a,D

Q

0

1-

2-

3-

4-

5-

6-

7 -

8-

9 -

10

5 10 20 20 10 5

100 - ; C=

***\ * 100 X .~~~~~~~~~~~~~~~10

~~~~~~~~~~~~0 55

NO- concentration (pJM)

20

20

*1 I I

A S 0 N D J F M A M J J A S 0 N DFIG. 1. Depth distribution and seasonal variation of in situ N03- concentrations in the water column and the littoral sediment of Lake

Hampen (data are based on the mean profile of six cores; standard deviation <40%).

The rest was located in the underlying sandy part of the rootzone. The definition of upper and lower root zones wasfurther supported by the slightly inhibitory effect of illumi-nation on the upper root zone denitrification (Fig. 3). Thiseffect was not observed in the lower root zone.

Seasonal variation of NO3-stimulated denitrification. TheNO3- addition to the sediment gave no stimulation ofdenitrification in the winter, irrespective of the light condi-tions (Fig. 4); low and almost similar rates were recorded inthe control and the NO3 -amended cores (Fig. 3 and 4; notescale change). In contrast, the stimulation by NO3 was verysignificant in the summer. Here the activities in the darkincreased up to 20- to 40-fold in the sediment from a depth of0 to 3 cm. For the incubations in the light, the stimulationwas somewhat lower: in the upper 3 cm the denitrification

1 '

2-

3.

I 4-

8

9.

J F M A M J J A S 0 N D

was up to 5- to 10-fold higher than under in situ conditions.For the lower part of the root zone, which had little seasonalchange of the NO3 concentrations in situ, we observed littleor no stimulation of the denitrification by NO3- amendment.

DISCUSSIONOur data demonstrate the presence of two denitrification

zones in the plant-covered sediment: one in the uppermostdebris layer, which was detectable throughout the year butwas strongly affected by light-dark cycles, and one in theunderlying sediment, which was slightly affected by light ata depth of 1 to 3 cm (upper root zone) but not at 3 to 10 cm(lower root zone).The temperature variation obviously plays an important

role in the seasonal pattern of denitrification. In the summer,

J F M A M J J A S O N D

FIG. 2. Depth distribution and seasonal variation of in situ denitrification activity in the L. uniflora-colonized sediment under light (A) anddark (B) conditions.

Light

0.1

* 005

-05/

APPL. ENVIRON. MICROBIOL.

DENITRIFICATION IN LAKE SEDIMENT 1177

60

50-

z

zE

c0

0

0

._co._

._

J F M A M J J A S O N D J F M A M J J A S 0 N D

FIG. 3. Seasonal variation of overall in situ denitrification in the L. uniflora-colonized sediment under light (A) and dark (B) conditions.Surface zone, 0 to 1 cm; upper root zone, 1 to 3 cm; lower root zone, 3 to 10 cm.

for instance, the denitrification was high, as would beexpected from a stimulation by the temperature alone.However, more elaborate diel and seasonal control patternscould be expected from the variations of light exposure andavailability of NO3 and organic substrates. The possible

400 -

350-

E

E

a0

.O-0

0

300-

250 -

200

150-

100-

50-

J F M A M J J A S O N

influence of such factors on the sediment denitrification isdiscussed below.

Denitrification in the surface zone. OUir measurementsindicate a strong diel variation of the denitrification in littorallake sediments. A similar pattern has recently been reported

D J F M A M J J A S O N D

FIG. 4. Seasonal variation of overall N03--stimulated denitrification in the L. uniflora-colonized sediment under light (A) and dark (B)conditions. Surface zone, 0 to 1 cm; upper root zone, 1 to 3 cm; lower root zone, 3 to 10 cm.

Light

Surfacezone

Upper,,.rroot'zone

Lowerrootzone

VOL. 51, 1986

1178 CHRISTENSEN AND S0RENSEN

for sediments from shallow marine waters (4). The stronginfluence of light on surface denitrification indicated that theactivity was located in the uppermost surface layer, in whichalgal photosynthesis was likely to control the 02 profile andits depth of penetration. The illumination of sediments maycause both a significant accumulation and a deeper penetra-tion of the 02. This has often been demonstrated forintertidal marine sediments, in which diatoms and othermicroalgae are abundant in the benthos (see, e.g., reference15). Except from the early summer, when the highest con-centrations were observed at a depth of 1 to 2 cm, themaximum NO3- accumulation was always found within theuppermost 1 cm (Fig. 1). This indicated that the surfacedenitrification was primarily supported by NO3- from thenitrification occurring close to the sediment-water interface.A control by the NO3- availability was apparent for both

the surface denitrification and the activity in the upper partof the root zone. Both layers were markedly affected byadditions of NO3- in the summer (Fig. 3 and 4). A maximumof stimulation was thus obtained at the onset of the growthseason (May-June), when a sudden decrease in the NO3-concentrations occurred in situ (Fig. 1). The apparent NO3-limitation of denitrification could be related to the highassimilatory demands for inorganic nitrogen in both theplants and the benthic microalgae at this time of year. Afterthe growth season, the N03- concentrations were againincreasing In situ, and the stimulation by an excess of NO3-could no longer be observed in the cores.

Denitrificatiohl in the root zone. The deep denitrificationzone, which developed in the summer, was clearly associ-ated with the root zone of L. uniflora. The primary produc-tivity of L. uniflora is significant only in the summer (18), andboth the seasonal pattern and the depth distribution there-fore strongly indicate a regulation of the denitrification bythe release of organic substrate from the roots.

Interactions between the plant and the root zonedenitrification were also indicated from the accumulation ofNO3- at a depth of 1 to 2 cm in the summer, which couldonly be explained by the presence of 02 and a secondaryzone of nitrification in this part of the root zone. Thenitrification must be supported by the release of 02 from theroots when the plant is undergoing active photosynthesis(17). The upper and lower denitrification zones appearing inthe summer may be separated by this intermediate zone ofnitrification. Such i zonal model, incorporating a deepnitrification zone, has been proposed for marine sediments inwhich the burrows bf the benthic fauna provide a route of 02transport to the deeper layers (11). The NO3 from a deepnitrification zone may support the denitrification to consid-erable depths in the sediment.A significant light-dependent release of organic com-

pounds has also been reported for L. uniflora roots (20). Asimultaneous release of both the 02 and the organic sub-strate may therefore indicate a complicated diel control ofroot zone denitrification. During the day, for instance, astimulatory effect of the organic compounds could well becounteracted by the inhibitory effect of 02. For the upperpart of the root zone, our data indicated higher activities inthe dark than in the light, but in the lower part of the rootzone, in which the organic substrate seems a major limitingfactor of denitrification, we found no such differences in theactivity for dark and light conditions. It was impossible tosay, however, whether the releases of 02 and organic carbonwere both of minor significance during the incubations orwhether their effects on the denitrification were truly antag-onistic.

As noted above, an apparent NO3- limitation was foundfor the denitrification in the upper part of the root zone. TheNO3--amended cores demonstrated a high potential fordenitrification in the summer and revealed a significantpopulation of denitrifying bacteria in the root zone. It wasinteresting that the light had a pronounced effect on theupper rdot zone denitrification in the NO3--amended sedi-ment (Fig. 4). Such a marked inhibition by the light wasprobably due to a release of 02 from the roots and wouldonly be expected if the population of denitrifiers were closelyassociated with the roots and therefore highly susceptible toplant metabolism.The present study clearly demonstrates the important role

of root zone denitrification in the littoral sediment of LakeHampen. On an annual basis, the root zone accounted forabout 70% of the total activity in the light and about 50% ofthe total in the dark. Comparable but nonvegetated littoralsediments showed low denitrification activity, and the activ-ities were restricted to the surface layer (P. B. Christensen,unpublished data). To determine nitrogen losses bydenitrification in lakes, it is therefore important to incorpo-rate direct measurements of the activity in the littoralplant-sediment community and to consider both the tempo-ral and spatial variations of the process.

ACKNOWLEDGMENT

The skillful technical assistance of Else B. Frentz is gratefullyacknowledged.

LITERATURE CITED1. Andersen, F. 0. 1981. Oxygen and nitrate respiration in a reed

swamp sediment from an eutrophic lake. Holarct. Ecol.4:66-72.

2. Andersen, J. M. 1974. Nitrogen and phosphorus budgets and therole of sediments in six shallow Danish Lakes. Arch. Hydrobiol.74:524-550.

3. Andersen, J. M. 1977. Rates of denitrification of undisturbedsediment from six lakes as a function of nitrate concentration,oxygen and temperature. Arch. Hydrobiol. 80:147-159.

4. Andersen, T. K., M. H. Jensen, and J. S0rensen. 1984. Diurnalvariation of nitrogen cycling in coastal marine sediments. I.Denitrification. Mar. Biol. (Berlin) 83:171-176.

5. Armstrong, F. A. J., C. R. Stearns, and J. D. H. Strickland.1967. The measurement of upwelling and subsequent biologicalprocesses by means of the Technicon Autoanalyzer and associ-ated equipment. Deep-Sea Res. 14:381-389.

6. Armstrong, W. 1964. Oxygen diffusion from the roots of someBritish bog plants. Nature (London) 204:801-802.

7. Brylinsky, M. 1977. Release of dissolved organic matter bysome marine macrophytes. Mar. Biol. (Berlin) 39:213-220.

8. Capone, D. G., P. A. Penhale, R. S. Oremland, and B. F. Taylor.1979. Relationship between productivity and N2 (C2H2) fixationin a Thallassia testudinum community. Limnol. Oceanogr.24:117-125.

9. Chan, Y. K., and N. E. R. Campbell. 1980. Denitrification inlake 227 during summer stratification. Can. J. Fish. Aquat. Sci.37:506-512.

10. Chan, Y. K., and R. Knowles. 1979. Measurement of denitrifica-tion in two freshwater sediments by an in situ acetylene inhibi-tion method. Appl. Environ. Microbiol. 37:1067-1072.

11. Grundmanis, V., and J. W. Murray. 1977. Nitrification anddenitrification in marine sediments from Puget Sound. Limnol.Oceanogr. 22:804-813.

12. Hough, R. A., and R. G. Wetzel. 1975. The release of dissolvedorganic carbon from submersed aquatic macrophytes: diel,seasonal, and community relationships. Verh. Int. Ver. Theor.Angew. Limnol. 19:939-948.

13. lizumi, H., A. Hattori, and C. P. McRoy. 1980. Nitrate andnitrite in interstitial waters of eelgrass beds in relation to the

APPL. ENVIRON. MICROBIOL.

DENITRIFICATION IN LAKE SEDIMENT

rhizosphere. J. Exp. Mar. Biol. Ecol. 47:191-201.14. Kemp, W. M., R. L. Wetzel, W. R. Boynton, C. F. D'Elia, and

J. C. Stevenson. 1982. Nitrogen cycling and estuarine interfaces:some current concepts and research directions, p. 209-230. InV. S. Kennedy (ed.), Estuarine comparisons. Academic Press,Inc., New York.

15. Revsbech, N. P., J. S0rensen, T. H. Blackburn, and J. P.Lomholt. 1980. Distribution of oxygen in marine sedimentsmeasured with microelectrodes. Limnol. Oceanogr. 25:403-411.

16. Sand-Jensen, K. 1978. Metabolic adaption and vertical zonationof Littorella uniflora (L.) Aschers. and Isoetes lacustris L.Aquat. Bot. 4:1-10.

17. Sand-Jensen, K., C. Prahl, and H. Stokholm. 1982. Oxygenrelease from roots of submerged aquatic macrophytes. Oikos38:349-354.

18. Sand-Jensen, K., and M. S0ndergaard. 1978. Growth and pro-

duction of isoetids in oligotrophic Lake Kalgaard, Denmark.Verh. Int. Ver. Theor. Angew. Limnol. 20:659-666.

19. Sherr, B. F., and W. J. Payne. 1977. Effect of the Spartinaalterniflora root-rhizome system on salt marsh soil denitrifyingbacteria. Appl. Environ. Microbiol. 35:724-728.

20. S0ndergaard, M. 1983. Heterotrophic utilization and decompo-sition of extracellular carbon released by the aquaticangiosperm Littorella uniflora (L.) Aschers. Aquat. Bot.16:59-73.

21. S0rensen, J. 1978. Denitrification rates in a marine sediment asmeasured by the acetylene inhibition technique. Appl. Environ.Microbiol. 36:139-143.

22. Tiren, T., J. Thorin, and H. Nommik. 1976. Denitrificationmeasurements in lakes. Acta Agric. Scand. 26:175-184.

23. Weiss, R. F., and B. A. Price. 1980. Nitrous oxide solubility inwater and sea water. Mar. Chem. 8:347-359.

VOL. 51, 1986 1179