Microb Ecol (1989) 17:143-157
MICROBIAL ECOLOGY 9 Springer-VerlagNew York Inc. 1989
Nitrate Requirement for Acetylene Inhibition of Nitrous Oxide Reduction in Marine Sediments Jennifer M. Siater ~ and Douglas G. Capone 2 ~MarineSciencesResearch Center, SUNY, Stony Brook, New York 11794, USA; and 2Centerfor Environmental and Estuarine Studies, Chesapeake BiologicalLaboratory, Solornons, Maryland 20688-0038, USA
Abstract.
The inhibition of nitrous oxide (N20) reduction by acetylene (C2H2) in saltmarsh sediment was temporary; we investigated this phenomenon and possible causes. The reduction of N20 in the presence of C2Hz was biological. N20 consumption in the presence of C2H2 began when nitrate concentration became very low. The time course o f N 20 consumption after periods of N20 accumulation was unaffected by initial nitrate concentrations between 16 and 200 #M, or CzH2 concentrations between 10 and 100% o f the gas phase. Sulfide had no effect on the kinetics of NzO reduction in the presence o f C2Hz. In more dilute slurries of saltmarsh sediments and in estuarine sediment, N 20 persisted in the presence of C2H2 unless sufficient organic carbon was added to deplete nitrate. In saltmarsh sediments, the rate of N20 consumption in the presence of C2H2 was not changed by preincubation with C2H2. Initial positive rates of N20 production in the presence o f C2H2 occurred only when the block was apparently effective (i.e., at nitrate concentrations greater than about 5-10 uM) and appeared to represent a valid estimate of denitrification. Conversely, and in agreement with previous studies, concentrations o f N O 3- below these levels resulted in reduced efficiency of C2H 2 blockage of N20 reductase.
Introduction Denitrification, the anaerobic respiratory reduction of nitrate to the gaseous end products N2 and N20, represents a loss o f combined nitrogen. Because nitrogen is often a limiting nutrient in marine systems, the magnitude o f this loss is o f great interest. Until recently, comparatively few direct measurements of denitrification rates have been made [11]. The mass balance technique sometimes employed in lakes [3, 19] requires detailed N and P cycling information that is rarely available, and most methods used require some kind of enclosure and incubation o f samples. Loss of nitrate from an incubation cannot necessarily be equated with denitrification, as other processes may contribute to the observed nitrate loss. For instance, conditions favoring denitrification might also favor
144
J . M . Slater and D. G. Capone
dissimilatory nitrate reduction to ammonium [13, 16, 17]. This objection applies also to assessments of denitrification rates from ambient nitrate concentration profiles [35]. Direct chromatographic measurement of nitrogen production is appealing and has been used [14, 25], but leakage of atmospheric nitrogen into incubation vessels and syringes may cause artifactual results [24]. The only widely available nitrogen isotope, ~SN, is tedious and expensive to analyze and, more important, the detection methods (mass spectrometry or emission spectrometry) are quite insensitive, requiring long incubation and high enrichment. Nitrogen-15 techniques for denitrification often require addition of extremely high nitrate concentrations [ 12, 29], although a more recent modification by Nishio et al. [21] avoids this problem. However, this procedure may only assay denitrification near the sediment-water interface. The only radioactive isotope of nitrogen that is useful, t3N [10, 28], has a half-life of a mere 10 min, and can only be used at a cyclotron facility. It is therefore of limited availability. An assay method that has promise is the C2H 2 blockage procedure. Acetylene has been shown to prevent reduction of N20 to N2, the last step in the denitrification pathway, in culture [4, 39], in soils [40], and in fresh water [8] and marine sediments [13, 29]. N20 is easily measured with high sensitivity by electron-capture gas chromatography; measurements are possible with short incubations and at realistic nitrate concentrations. The inhibition of N20 reductase, however, sometimes fails, particularly when nitrate concentration is very low [15, 16, 22, 28]. In other cases, the inhibition is apparently only temporary [22, 26, 36, 38]. Sulfide, which can directly inhibit N20 reductase [31 ], has also been shown to reduce the efficacy of the C2H2 block [34]. The ease, speed, and sensitivity of the gas chromatographic assay for N20 make the C2H2 blockage method an attractive choice for assessing in situ denitrification. Its applicability, according to our own results, varies with sediment type. We attempted to establish the utility of the technique in a saltmarsh sediment, where N20 accumulates only temporarily in the presence of C2H2 [26]. We also investigated factors involved in the N20 disappearance, partly by comparison with another system in which the C2H2 block had always worked well. Materials and M e t h o d s
Sites, Sampling, and Setup Procedures Saltmarsh sediment was obtained from Flax Pond, Stony Brook, New York, from a low-marsh, intertidal zone, among Spartina alterniflora plants on the edge of a mud panne. Sediment cores were taken in 2.5 cm diameter aluminum core tubes. Organic content of the sediments was typically about 5% (range from 3 to 7% loss on ignition at 450~ The top 10 cm of the cores was homogenized in a Waring blender with deoxygenated, filtered (0.45 t*M Gelman Metricel) seawater in the ratio 2:1 (seawater: sediment, vol : vol). Root/rhizome material was removed by forcing the slurry through a 2 mm sieve. The slurry was dispensed in 50 ml portions (measured by pipette) into 125 ml Erlenmeyer flasks, which were immediately stoppered. Throughout setup, the sediment was gassed with O2-free N2, and each flask was gassed for 2 rain after stoppering. Estuarine sediment was obtained from Great South Bay, NY, from a subtidal site offPatchogue.
Nitrate for Acetylene Inhibition of N20
145
Water depth was about 1-1.5 m. Sediment was sandy, with sparse eelgrass patches, with an organic content of about one-tenth that of Flax Pond (0.5% loss on ignition). The top 8 cm was taken directly into 50 ml syringes with the end opposite the plunger removed. The syringe cores were returned to the laboratory within 2 hours. This sand was too coarse to be slurried. The sand from all the syringes was mixed together, and 50 ml volumes were dispensed into 125 ml Erlenmeyer flasks. This operation was performed in a glove-bag filled with O2-free N2. Additions of nitrate and glucose were made in 15 ml of filtered overlying water, which had been bubbled with O2-free Na for 30 min. Flasks were stoppered within the glove-bag and then subsequently gassed with N2 for 2 rain. Additions of nitrate, organic carbon, and S2- were made as KNO3, glucose, and Na2S, respectively. Amendments to the slurries were volumes of 0.5 ml. Additions to the sandy sediment were made in filtered seawater in a volume of 15 ml to ensure rapid penetration 9 Nitrate additions were to a final concentration of either 100 (initial experiments) or 200 lzM, unless otherwise noted.
Denitrification Measurements Denitrification determinations were performed on the sediment slurries using the C2H2 blockage [4, 29, 39, 40] or N20 reductase [20] procedure. For C2H2block assays, C2H2 was added to the gas phase of the incubations to a concentration usually of 15-20%, and N20 concentration in the gas phase atmosphere was monitored with time. Nitrous oxide reductase assays were performed by adding about 100 ppm (vol:vol) N20 to the gas phase of the incubation flasks and monitoring changes in its concentration with time. Concentrations of NzO in the gas phase were measured on a Perkin-Elmer Sigma 2B gas throat atograph with a 63Ni electron-capture detector. Samples o f 100 ~1 were withdrawn periodically om the gas phase of the incubations and injected onto a 2 m Porapak Q column with 3 m m external diameter. Column temperature was 60~ detector temperature 300"C, and the carrier gas was 95% argon/5% methane (P5 mix) at a flow rate of 34 ml/min. A six-way valve allowed backflush to vent in order to prevent C2H2 contamination of the detector in those flasks that contained C2H2. The detector was calibrated before and after every time point with gas standards close to the COncentrations in the incubations. Concentration of N20 estimated in the gas phase was corrected for that in solution. C2H~ and C2H 4 concentrations were determined on separate gas samples by flame-ionization gas chromatography.
Nitrate, Nitrite, and Ammonium Determinations Replicate flasks were periodically sacrificed for determinations of nitrate, nitrite, and ammonium. Pore waters were extracted by centrifugation at 3,000 rpm for 20 min in Delrin pore-water bottles (B. Brinkhuis, patent pending) and immediately frozen. Analyses were performed by standard methods [33] adapted for a Technicon autoanalyzer [37]. Sulfide was removed from the samples before nutrient determinations by treatment with CdCI2.
PH Determinations PI-t changes over the time course of the incubations were recorded for the saltmarsh sediment on Several occasions. One replicate flask was sealed with a stopper, through which a pH probe (a Sensorex combination gel electrode) had been inserted. No significant changes in pH were noted Over the time course of the experiments. 9 Unless otherwise noted, values given are means of three replicates with standard error. Significance level was p = 0.05.
146
J . M . Slater a n d D. G. C a p o n e
500
o
,11
x
0.85), provided oxygen concentration was below 0.05%. The increase in N20 which continued for 10-27 hours was followed by a period of 4-20 hours of constant N20 concentration. This was then followed by a rapid decrease of N20, which always disappeared by 120 hours. Initial rates of N20 production in slurries amended with 100 ~zM nitrate varied between 5.4 (_+0.81) and 35.6 (+1.3) nmole N20 g-~h -1. Maximum N20 accumulation was between 21.4 (_+2.6) and 489 (_+ 15.4) nmole g-), and represented 10-87% of the initially available nitrate-N. The greatest proportion of added nitrate converted to N2 occurred in samples collected in the late fall, winter, and spring, and the lowest in late summer (data not shown). The change in concentrations of nitrate, nitrite, and ammonium and the production of N20 throughout an incubation in the presence of CzH2 is shown for one experiment in Figure 1. These trends were consistently observed in several similar experiments. NzO did not begin to disappear from any of these incubations until nitrate concentration fell, in this case to below 5.3 #M (Fig. 1) and to below 3.6 #M in another experiment. In the experiment shown in Fig. 1, several flasks were left without added nitrate in order to assess by difference the amount of ammonium produced by dissimilatory reduction of
147
Nitrate for Acetylene Inhibition of N2O 800
600 Ul
~ 400
200
0
20
40
60
TIME (h)
Fig. 2. Effectsof initial nitrate concentration on nitrous oxide production in the presence of 15% acetylene in saltmarsh sediment slurries; ([~) 23 /~M (ambient); (l) 50 /~M; (0) 100 uM; (A) 200 uM. Means of three replicates _+ standard error.
nitrate. T h e nitrogen balance at 24 hours was as follows: nitrate-N used, 9.2 ug-at per flask; a m m o n i u m apparently generated f r o m added nitrate, 4.7 #g-at Per flask; N 2 0 - N accumulated in the presence o f C2H2, 3.5 #g-at per flask. Thus, 89% o f the N was accounted for. Various initial nitrate concentrations, although affecting both the initial rate of N 2 0 p r o d u c t i o n and its total amount, did not change the characteristics o f the time course (Fig. 2). T h e r e was no difference in the time course or magnitude of N 2 0 accumulation a m o n g incubations in several experiments using C2H2 Concentrations between 10 and 100% (data not shown). In several experiments, nitrous oxide disappearance in the presence o f Call2 was completely prevented, relative to controls, by addition o f 1% (vol: vol) formalin at the time o f maximum N~O accumulation. C2H2 persisted at steady levels through all incubations. There was no evidence for the c o n s u m p t i o n o f C2H2 as a factor contribUting to failure o f the block. In slurries diluted to greater than a 10:1 ratio, N 2 0 persisted for at least 340 hours. Initial rates and m a x i m u m levels o f NzO (normalized to sediment dry Weight) were not significantly different from those o f controls (i.e., 2:1 slurries) until dilutions reached 50:1, when initial long lags in N 2 0 p r o d u c t i o n were apparent, followed by eventual very high production. Figure 3 shows the perSistence o f N 2 0 in a 20:1 slurry ( s e a w a t e r : s e d i m e n t , v o l : v o l ) . In another experiment, N 2 0 concentration was 490 n m o l g-~ in the 20:1 slurry after 7 Clays with a concurrent nitrate concentration o f 27 #M. In the 2:1 slurry, N 2 0 had disappeared by 7 days and nitrate was undetectable. These m o r e diluted slurries were also used to investigate a possible role o f Sulfide in N 2 0 disappearance. A m o n g the treatments o f 20:1 slurries (without
148
J . M . Slater and D. G. Capone
350
--
__
~176 f 300
0
E 250
--~ 200 150 nt-
100 5O
0
I
I
I
20
40
60
80
I
I
I
100
120
140
160
TIME (h)
Fig. 3. Effects o f dilution and of added sulfide on nitrous oxide reduction in saltmarsh sediment in the presence o f acetylene. (E3) 2:1 slurry; ~1) 20:1 slurry; (O) 20:1 slurry + 0.4 m M sulfide; (Zx) 20:1 slurry + 4 m M sulfide. Bars represent the standard error o f the mean.
sulfide, with 0.4 mM, and with 4 mM added sulfide) there were no statistically significant differences in initial rates, maximum production, or N20 persistence (Fig. 3), although the slope of the 4 mM treatment did appear to be reduced and the mean overall production may have been slightly greater. The rate of N20 production in the presence of C2H2 was compared with the rate of disappearance of added N20 in the absence of C2H2 (Fig. 4), an alternative method of assessing denitrification [20]. Nitrate was also added to these incubations (200 /~M); initial added N20 concentration was about 100 ppm (vol : vol) in the gas phase. N20 consumption in the absence of acetylene occurred after a brief period ( ~ 2 hours) of initial equilibration with the liquid phase, followed by a period (9 hours in this case) during which N20 in the gas phase remained constant. Rates were calculated after this lag period. With an initial concentration of N20 of 104 nmole g-l, the rate of accumulation of N20 in the presence of C2H2 (7.1 _ 0.63 nmole g-t h - l , r 2 = 0.96) and the rate of disappearance of N20 in the absence of C2H2 (5.7 +__0.01 nmole g-Zh-'; r 2 = 0.95) were not significantly different from each other. The rate of disappearance of added N20 (100 ppm) in the presence and absence of C2H2 was measured in another experiment in which no nitrate was added to the sediment slurry. Only two replicates were used, the others being treated with 1% formalin. Formalin addition prevented any change in N20 concentration, either with or without C2H2, except for an initial (0-2 hours) drop of 12% in headspace N20, presumably as a result of equilibration. In contrast to the experiment described above, N20 disappearance without C2H2 proceeded without a lag and yielded a rate of 8.8 + 0.01 nmole g - ' h -] (r 2
Nitrate for Acetylene Inhibition of N20
149
180 160 140 120 100 ~ 8O 6O 4O 2O OI 0
=
i 20
~
T
i
40
6o
Time (h)
Fig. 4. Time course of N20 concentration in sediment samples incubated with 200 #M nitrate and 20% C2Hz(n), or with 100 ppm N:O, 200 uM nitrate and no CzH~(O). Bars represent standard error of the mean. .0.93) over the first 10 hours. In the presence o f C2H2, N20 concentration first increased slightly, but after 17 hours it began a linear decrease. Rate o f disappearance from 17 to 30 hours was 8.6 -+ 0.70 nmole g-~h -~ (r 2 = 0.99). There is no significant difference between these rates. The effects o f preincubation o f the sediment slurry with C2H2 were assessed in two experiments. In the first e x p e r i m e n t (Table 1), incubations were carried out without C2H2, and with additions o f C2H2 at 0 and 24 hours. In all cases, N20 was added at 24 hours. After 24 hours incubation, N 2 0 reduction occurred in all flasks, including those with C2H2 added at either 0 time or at 24 hours. There was no significant difference between the incubations with C2H2 added at 0 time or 24 hours, but each o f these rates was about 70% o f the rate in the C2H2-free control (p < 0.05), possibly because o f slightly higher concentrations of N20 in control (C2H2-free) flasks. In a second preincubation experiment, subsamples o f slurry were taken from parent incubations, either with or without C2H2, at 10 hours (while N 2 0 was Still accumulating in the presence o f C2H2), at 13 hours (when N 2 0 concentration had leveled off), and at 16 hours (when N 2 0 had begun to disappear from the parent incubation in the presence o f C2H2 (see Fig. 5A). T h e subsamples Were then reincubated with addition o f 100 p p m N20, with and without C2H2. The disappearance o f N 2 0 in these subsamples is shown in Fig. 5B, C, and D. At 10 hours, C2H2 was effective in blocking N 2 0 reduction from samples either Preincubated with or without C2H2 (Fig. 5B). At 13 hours, C2H2 was only 13artially effective in preventing N 2 0 c o n s u m p t i o n (Fig. 5C). By 16 hours, C2H2 Was essentially ineffective in blocking N 2 0 reduction whether the preincubation Was with or without C2H2 (Fig. 5D).
150
J.M. Slater and D. G. Capone
Table 1. Effectsof 24 hour preincubation with acetylene on nitrous oxide reduction by saltmarsh sediment slurry~ Time of C2H2 addition
Rate of N20 reduction when added at 24 hours (nmole g-~h-~)
Initial N20 concentration (nmole g ~)
0h 24 h No C2H2
12.6 _+ 0.170 13.2 +_ 0.610 18.1 + 0.260
89.1 _+ 0 92.1 -+ 4.78 127 + 1.45
Values are means of two replicates _+ standard error
Estuarine Sediment Previous e x p e r i m e n t s at this site showed that N 2 0 a c c u m u l a t e d and persisted for up to 2 weeks in the presence o f C2H 2 [27]. T h e concentrations o f nitrate, nitrite, a n d a m m o n i u m , as well as the a c c u m u l a t i o n o f N 2 0 , were followed in two experiments. In the first experiment, nitrate concentration n e v e r fell below 7.2/~M, and N 2 0 persisted for at least 121 hours (data not shown). These sediments h a v e low organic content ( < 1% loss on ignition); therefore, in one experiment, glucose was a d d e d to a concentration o f 20 raM. Figure 6 shows the variations in NO3- a n d N 2 0 concentrations, with (B) and without (A) glucose. N 2 0 began to decrease in the incubations with glucose at 77 hours, when nitrate concentration was 3.2 ~zM, a n d was zero by 7 days. In the replicates without a d d e d glucose, N~O persisted up to at least 168 hours, a n d nitrate concentration was effectively constant f r o m 20 h o u r s at 7.49 (_+0.38) uM.
Discussion In our sediments, we n e v e r o b s e r v e d 100% c o n v e r s i o n o f nitrate to N20 and s o m e t i m e s it was as low as 10%. Similar results h a v e been reported by O r e m l a n d et al. [22] in intertidal sediments and b y C a p o n e a n d T a y l o r [7] in tropical seagrass sediments. This m a y be due to i n c o m p l e t e n e s s o f the CEH 2 block or to other processes using nitrate, or both. One possibility is the dissimilatorY reduction o f nitrate to a m m o n i u m . In m a r i n e sediments, the p r o p o r t i o n of nitrate being reduced to a m m o n i u m rather than to N2 has been s h o w n to be as m u c h as 70% o f the total [18, 30], a n d to increase at low redox potential [6] and at low nitrate concentration [17]. In the saltmarsh sediment, 89% o f the a d d e d nitrate was accounted for in one experiment, 51% going to a m m o n i u m and 38% to N20. T h e l l % u n a c c o u n t e d for m a y simply h a v e been due to e x p e r i m e n t a l error, to the fact that we m e a s u r e d only free a m m o n i u m , to other processes using nitrate (e.g., assimilation), or to incompleteness o f the C2H2 block. T h o u g h the C2H2 block m e t h o d has been used with a p p a r e n t success by Yoshinari et al. in soils [40] and by Sorensen in m a r i n e s e d i m e n t s [29], and
Nitrate for Acetylene Inhibition of NzO
151
4o
ao,
z
1
0
0 10
I
I
~
I
1
12
I
_t
14
I
1
16
1
t
18
,
9
20
22
120 - 100
~ 8C
!oo
!
20 ~
0
.a-
2
9
I.
i
4
0
I
L ,
I
2 4 TIME (h)
0
2
4
6
Fig. S. Effectsof preincubation of saltmarsh sediment; 200 #M nitrate added at zero time. (A) Preincubationwith (11)and without ({3)acetylene. Panels 8, C, and D show nitrous oxide reduction In reincubations taken from the parent preincubations at the points marked in A: preincubation and reincubation with acetylene (ml);preincubation without and reincubation with acetylene (0); l~reincubationwith and reincubation without acetylene ([3);preincubation and reincubation without acetylene (O). Bars represent standard error of the mean.
has been validated by c o m p a r i s o n with 13N in soils [28] and with ~SN in sediments [27, 29], some inconsistencies have been observed [23]. Y e o m a n s and Beauchamp [38], working with soils, noted that N 2 0 began to disappear from incubations with 1 and 10% C2H2 after 100 and 160 hours, respectively. Van Raalte and Patriqu~n [361 obtained inconsistent results with saltmarsh sediment. In one experiment, accumulated N 2 0 disappeared between 4 and 8 Clays in the presence o f 10% C2Hz; in a n o t h e r experiment, the N 2 0 persisted. Our own previous investigations [26] indicated that in Flax P o n d saltmarsh Sediment, N~O accumulated only temporarily in the presence o f C2H2; time
152
J.M. Slater and D. G. Capone
A
240
~0
200 160
20
120 '3
30
0,~,~ 10 !
40 o
W
MICROBIAL ECOLOGY 9 Springer-VerlagNew York Inc. 1989
Nitrate Requirement for Acetylene Inhibition of Nitrous Oxide Reduction in Marine Sediments Jennifer M. Siater ~ and Douglas G. Capone 2 ~MarineSciencesResearch Center, SUNY, Stony Brook, New York 11794, USA; and 2Centerfor Environmental and Estuarine Studies, Chesapeake BiologicalLaboratory, Solornons, Maryland 20688-0038, USA
Abstract.
The inhibition of nitrous oxide (N20) reduction by acetylene (C2H2) in saltmarsh sediment was temporary; we investigated this phenomenon and possible causes. The reduction of N20 in the presence of C2Hz was biological. N20 consumption in the presence of C2H2 began when nitrate concentration became very low. The time course o f N 20 consumption after periods of N20 accumulation was unaffected by initial nitrate concentrations between 16 and 200 #M, or CzH2 concentrations between 10 and 100% o f the gas phase. Sulfide had no effect on the kinetics of NzO reduction in the presence o f C2Hz. In more dilute slurries of saltmarsh sediments and in estuarine sediment, N 20 persisted in the presence of C2H2 unless sufficient organic carbon was added to deplete nitrate. In saltmarsh sediments, the rate of N20 consumption in the presence of C2H2 was not changed by preincubation with C2H2. Initial positive rates of N20 production in the presence o f C2H2 occurred only when the block was apparently effective (i.e., at nitrate concentrations greater than about 5-10 uM) and appeared to represent a valid estimate of denitrification. Conversely, and in agreement with previous studies, concentrations o f N O 3- below these levels resulted in reduced efficiency of C2H 2 blockage of N20 reductase.
Introduction Denitrification, the anaerobic respiratory reduction of nitrate to the gaseous end products N2 and N20, represents a loss o f combined nitrogen. Because nitrogen is often a limiting nutrient in marine systems, the magnitude o f this loss is o f great interest. Until recently, comparatively few direct measurements of denitrification rates have been made [11]. The mass balance technique sometimes employed in lakes [3, 19] requires detailed N and P cycling information that is rarely available, and most methods used require some kind of enclosure and incubation o f samples. Loss of nitrate from an incubation cannot necessarily be equated with denitrification, as other processes may contribute to the observed nitrate loss. For instance, conditions favoring denitrification might also favor
144
J . M . Slater and D. G. Capone
dissimilatory nitrate reduction to ammonium [13, 16, 17]. This objection applies also to assessments of denitrification rates from ambient nitrate concentration profiles [35]. Direct chromatographic measurement of nitrogen production is appealing and has been used [14, 25], but leakage of atmospheric nitrogen into incubation vessels and syringes may cause artifactual results [24]. The only widely available nitrogen isotope, ~SN, is tedious and expensive to analyze and, more important, the detection methods (mass spectrometry or emission spectrometry) are quite insensitive, requiring long incubation and high enrichment. Nitrogen-15 techniques for denitrification often require addition of extremely high nitrate concentrations [ 12, 29], although a more recent modification by Nishio et al. [21] avoids this problem. However, this procedure may only assay denitrification near the sediment-water interface. The only radioactive isotope of nitrogen that is useful, t3N [10, 28], has a half-life of a mere 10 min, and can only be used at a cyclotron facility. It is therefore of limited availability. An assay method that has promise is the C2H 2 blockage procedure. Acetylene has been shown to prevent reduction of N20 to N2, the last step in the denitrification pathway, in culture [4, 39], in soils [40], and in fresh water [8] and marine sediments [13, 29]. N20 is easily measured with high sensitivity by electron-capture gas chromatography; measurements are possible with short incubations and at realistic nitrate concentrations. The inhibition of N20 reductase, however, sometimes fails, particularly when nitrate concentration is very low [15, 16, 22, 28]. In other cases, the inhibition is apparently only temporary [22, 26, 36, 38]. Sulfide, which can directly inhibit N20 reductase [31 ], has also been shown to reduce the efficacy of the C2H2 block [34]. The ease, speed, and sensitivity of the gas chromatographic assay for N20 make the C2H2 blockage method an attractive choice for assessing in situ denitrification. Its applicability, according to our own results, varies with sediment type. We attempted to establish the utility of the technique in a saltmarsh sediment, where N20 accumulates only temporarily in the presence of C2H2 [26]. We also investigated factors involved in the N20 disappearance, partly by comparison with another system in which the C2H2 block had always worked well. Materials and M e t h o d s
Sites, Sampling, and Setup Procedures Saltmarsh sediment was obtained from Flax Pond, Stony Brook, New York, from a low-marsh, intertidal zone, among Spartina alterniflora plants on the edge of a mud panne. Sediment cores were taken in 2.5 cm diameter aluminum core tubes. Organic content of the sediments was typically about 5% (range from 3 to 7% loss on ignition at 450~ The top 10 cm of the cores was homogenized in a Waring blender with deoxygenated, filtered (0.45 t*M Gelman Metricel) seawater in the ratio 2:1 (seawater: sediment, vol : vol). Root/rhizome material was removed by forcing the slurry through a 2 mm sieve. The slurry was dispensed in 50 ml portions (measured by pipette) into 125 ml Erlenmeyer flasks, which were immediately stoppered. Throughout setup, the sediment was gassed with O2-free N2, and each flask was gassed for 2 rain after stoppering. Estuarine sediment was obtained from Great South Bay, NY, from a subtidal site offPatchogue.
Nitrate for Acetylene Inhibition of N20
145
Water depth was about 1-1.5 m. Sediment was sandy, with sparse eelgrass patches, with an organic content of about one-tenth that of Flax Pond (0.5% loss on ignition). The top 8 cm was taken directly into 50 ml syringes with the end opposite the plunger removed. The syringe cores were returned to the laboratory within 2 hours. This sand was too coarse to be slurried. The sand from all the syringes was mixed together, and 50 ml volumes were dispensed into 125 ml Erlenmeyer flasks. This operation was performed in a glove-bag filled with O2-free N2. Additions of nitrate and glucose were made in 15 ml of filtered overlying water, which had been bubbled with O2-free Na for 30 min. Flasks were stoppered within the glove-bag and then subsequently gassed with N2 for 2 rain. Additions of nitrate, organic carbon, and S2- were made as KNO3, glucose, and Na2S, respectively. Amendments to the slurries were volumes of 0.5 ml. Additions to the sandy sediment were made in filtered seawater in a volume of 15 ml to ensure rapid penetration 9 Nitrate additions were to a final concentration of either 100 (initial experiments) or 200 lzM, unless otherwise noted.
Denitrification Measurements Denitrification determinations were performed on the sediment slurries using the C2H2 blockage [4, 29, 39, 40] or N20 reductase [20] procedure. For C2H2block assays, C2H2 was added to the gas phase of the incubations to a concentration usually of 15-20%, and N20 concentration in the gas phase atmosphere was monitored with time. Nitrous oxide reductase assays were performed by adding about 100 ppm (vol:vol) N20 to the gas phase of the incubation flasks and monitoring changes in its concentration with time. Concentrations of NzO in the gas phase were measured on a Perkin-Elmer Sigma 2B gas throat atograph with a 63Ni electron-capture detector. Samples o f 100 ~1 were withdrawn periodically om the gas phase of the incubations and injected onto a 2 m Porapak Q column with 3 m m external diameter. Column temperature was 60~ detector temperature 300"C, and the carrier gas was 95% argon/5% methane (P5 mix) at a flow rate of 34 ml/min. A six-way valve allowed backflush to vent in order to prevent C2H2 contamination of the detector in those flasks that contained C2H2. The detector was calibrated before and after every time point with gas standards close to the COncentrations in the incubations. Concentration of N20 estimated in the gas phase was corrected for that in solution. C2H~ and C2H 4 concentrations were determined on separate gas samples by flame-ionization gas chromatography.
Nitrate, Nitrite, and Ammonium Determinations Replicate flasks were periodically sacrificed for determinations of nitrate, nitrite, and ammonium. Pore waters were extracted by centrifugation at 3,000 rpm for 20 min in Delrin pore-water bottles (B. Brinkhuis, patent pending) and immediately frozen. Analyses were performed by standard methods [33] adapted for a Technicon autoanalyzer [37]. Sulfide was removed from the samples before nutrient determinations by treatment with CdCI2.
PH Determinations PI-t changes over the time course of the incubations were recorded for the saltmarsh sediment on Several occasions. One replicate flask was sealed with a stopper, through which a pH probe (a Sensorex combination gel electrode) had been inserted. No significant changes in pH were noted Over the time course of the experiments. 9 Unless otherwise noted, values given are means of three replicates with standard error. Significance level was p = 0.05.
146
J . M . Slater a n d D. G. C a p o n e
500
o
,11
x
0.85), provided oxygen concentration was below 0.05%. The increase in N20 which continued for 10-27 hours was followed by a period of 4-20 hours of constant N20 concentration. This was then followed by a rapid decrease of N20, which always disappeared by 120 hours. Initial rates of N20 production in slurries amended with 100 ~zM nitrate varied between 5.4 (_+0.81) and 35.6 (+1.3) nmole N20 g-~h -1. Maximum N20 accumulation was between 21.4 (_+2.6) and 489 (_+ 15.4) nmole g-), and represented 10-87% of the initially available nitrate-N. The greatest proportion of added nitrate converted to N2 occurred in samples collected in the late fall, winter, and spring, and the lowest in late summer (data not shown). The change in concentrations of nitrate, nitrite, and ammonium and the production of N20 throughout an incubation in the presence of CzH2 is shown for one experiment in Figure 1. These trends were consistently observed in several similar experiments. NzO did not begin to disappear from any of these incubations until nitrate concentration fell, in this case to below 5.3 #M (Fig. 1) and to below 3.6 #M in another experiment. In the experiment shown in Fig. 1, several flasks were left without added nitrate in order to assess by difference the amount of ammonium produced by dissimilatory reduction of
147
Nitrate for Acetylene Inhibition of N2O 800
600 Ul
~ 400
200
0
20
40
60
TIME (h)
Fig. 2. Effectsof initial nitrate concentration on nitrous oxide production in the presence of 15% acetylene in saltmarsh sediment slurries; ([~) 23 /~M (ambient); (l) 50 /~M; (0) 100 uM; (A) 200 uM. Means of three replicates _+ standard error.
nitrate. T h e nitrogen balance at 24 hours was as follows: nitrate-N used, 9.2 ug-at per flask; a m m o n i u m apparently generated f r o m added nitrate, 4.7 #g-at Per flask; N 2 0 - N accumulated in the presence o f C2H2, 3.5 #g-at per flask. Thus, 89% o f the N was accounted for. Various initial nitrate concentrations, although affecting both the initial rate of N 2 0 p r o d u c t i o n and its total amount, did not change the characteristics o f the time course (Fig. 2). T h e r e was no difference in the time course or magnitude of N 2 0 accumulation a m o n g incubations in several experiments using C2H2 Concentrations between 10 and 100% (data not shown). In several experiments, nitrous oxide disappearance in the presence o f Call2 was completely prevented, relative to controls, by addition o f 1% (vol: vol) formalin at the time o f maximum N~O accumulation. C2H2 persisted at steady levels through all incubations. There was no evidence for the c o n s u m p t i o n o f C2H2 as a factor contribUting to failure o f the block. In slurries diluted to greater than a 10:1 ratio, N 2 0 persisted for at least 340 hours. Initial rates and m a x i m u m levels o f NzO (normalized to sediment dry Weight) were not significantly different from those o f controls (i.e., 2:1 slurries) until dilutions reached 50:1, when initial long lags in N 2 0 p r o d u c t i o n were apparent, followed by eventual very high production. Figure 3 shows the perSistence o f N 2 0 in a 20:1 slurry ( s e a w a t e r : s e d i m e n t , v o l : v o l ) . In another experiment, N 2 0 concentration was 490 n m o l g-~ in the 20:1 slurry after 7 Clays with a concurrent nitrate concentration o f 27 #M. In the 2:1 slurry, N 2 0 had disappeared by 7 days and nitrate was undetectable. These m o r e diluted slurries were also used to investigate a possible role o f Sulfide in N 2 0 disappearance. A m o n g the treatments o f 20:1 slurries (without
148
J . M . Slater and D. G. Capone
350
--
__
~176 f 300
0
E 250
--~ 200 150 nt-
100 5O
0
I
I
I
20
40
60
80
I
I
I
100
120
140
160
TIME (h)
Fig. 3. Effects o f dilution and of added sulfide on nitrous oxide reduction in saltmarsh sediment in the presence o f acetylene. (E3) 2:1 slurry; ~1) 20:1 slurry; (O) 20:1 slurry + 0.4 m M sulfide; (Zx) 20:1 slurry + 4 m M sulfide. Bars represent the standard error o f the mean.
sulfide, with 0.4 mM, and with 4 mM added sulfide) there were no statistically significant differences in initial rates, maximum production, or N20 persistence (Fig. 3), although the slope of the 4 mM treatment did appear to be reduced and the mean overall production may have been slightly greater. The rate of N20 production in the presence of C2H2 was compared with the rate of disappearance of added N20 in the absence of C2H2 (Fig. 4), an alternative method of assessing denitrification [20]. Nitrate was also added to these incubations (200 /~M); initial added N20 concentration was about 100 ppm (vol : vol) in the gas phase. N20 consumption in the absence of acetylene occurred after a brief period ( ~ 2 hours) of initial equilibration with the liquid phase, followed by a period (9 hours in this case) during which N20 in the gas phase remained constant. Rates were calculated after this lag period. With an initial concentration of N20 of 104 nmole g-l, the rate of accumulation of N20 in the presence of C2H2 (7.1 _ 0.63 nmole g-t h - l , r 2 = 0.96) and the rate of disappearance of N20 in the absence of C2H2 (5.7 +__0.01 nmole g-Zh-'; r 2 = 0.95) were not significantly different from each other. The rate of disappearance of added N20 (100 ppm) in the presence and absence of C2H2 was measured in another experiment in which no nitrate was added to the sediment slurry. Only two replicates were used, the others being treated with 1% formalin. Formalin addition prevented any change in N20 concentration, either with or without C2H2, except for an initial (0-2 hours) drop of 12% in headspace N20, presumably as a result of equilibration. In contrast to the experiment described above, N20 disappearance without C2H2 proceeded without a lag and yielded a rate of 8.8 + 0.01 nmole g - ' h -] (r 2
Nitrate for Acetylene Inhibition of N20
149
180 160 140 120 100 ~ 8O 6O 4O 2O OI 0
=
i 20
~
T
i
40
6o
Time (h)
Fig. 4. Time course of N20 concentration in sediment samples incubated with 200 #M nitrate and 20% C2Hz(n), or with 100 ppm N:O, 200 uM nitrate and no CzH~(O). Bars represent standard error of the mean. .0.93) over the first 10 hours. In the presence o f C2H2, N20 concentration first increased slightly, but after 17 hours it began a linear decrease. Rate o f disappearance from 17 to 30 hours was 8.6 -+ 0.70 nmole g-~h -~ (r 2 = 0.99). There is no significant difference between these rates. The effects o f preincubation o f the sediment slurry with C2H2 were assessed in two experiments. In the first e x p e r i m e n t (Table 1), incubations were carried out without C2H2, and with additions o f C2H2 at 0 and 24 hours. In all cases, N20 was added at 24 hours. After 24 hours incubation, N 2 0 reduction occurred in all flasks, including those with C2H2 added at either 0 time or at 24 hours. There was no significant difference between the incubations with C2H2 added at 0 time or 24 hours, but each o f these rates was about 70% o f the rate in the C2H2-free control (p < 0.05), possibly because o f slightly higher concentrations of N20 in control (C2H2-free) flasks. In a second preincubation experiment, subsamples o f slurry were taken from parent incubations, either with or without C2H2, at 10 hours (while N 2 0 was Still accumulating in the presence o f C2H2), at 13 hours (when N 2 0 concentration had leveled off), and at 16 hours (when N 2 0 had begun to disappear from the parent incubation in the presence o f C2H2 (see Fig. 5A). T h e subsamples Were then reincubated with addition o f 100 p p m N20, with and without C2H2. The disappearance o f N 2 0 in these subsamples is shown in Fig. 5B, C, and D. At 10 hours, C2H2 was effective in blocking N 2 0 reduction from samples either Preincubated with or without C2H2 (Fig. 5B). At 13 hours, C2H2 was only 13artially effective in preventing N 2 0 c o n s u m p t i o n (Fig. 5C). By 16 hours, C2H2 Was essentially ineffective in blocking N 2 0 reduction whether the preincubation Was with or without C2H2 (Fig. 5D).
150
J.M. Slater and D. G. Capone
Table 1. Effectsof 24 hour preincubation with acetylene on nitrous oxide reduction by saltmarsh sediment slurry~ Time of C2H2 addition
Rate of N20 reduction when added at 24 hours (nmole g-~h-~)
Initial N20 concentration (nmole g ~)
0h 24 h No C2H2
12.6 _+ 0.170 13.2 +_ 0.610 18.1 + 0.260
89.1 _+ 0 92.1 -+ 4.78 127 + 1.45
Values are means of two replicates _+ standard error
Estuarine Sediment Previous e x p e r i m e n t s at this site showed that N 2 0 a c c u m u l a t e d and persisted for up to 2 weeks in the presence o f C2H 2 [27]. T h e concentrations o f nitrate, nitrite, a n d a m m o n i u m , as well as the a c c u m u l a t i o n o f N 2 0 , were followed in two experiments. In the first experiment, nitrate concentration n e v e r fell below 7.2/~M, and N 2 0 persisted for at least 121 hours (data not shown). These sediments h a v e low organic content ( < 1% loss on ignition); therefore, in one experiment, glucose was a d d e d to a concentration o f 20 raM. Figure 6 shows the variations in NO3- a n d N 2 0 concentrations, with (B) and without (A) glucose. N 2 0 began to decrease in the incubations with glucose at 77 hours, when nitrate concentration was 3.2 ~zM, a n d was zero by 7 days. In the replicates without a d d e d glucose, N~O persisted up to at least 168 hours, a n d nitrate concentration was effectively constant f r o m 20 h o u r s at 7.49 (_+0.38) uM.
Discussion In our sediments, we n e v e r o b s e r v e d 100% c o n v e r s i o n o f nitrate to N20 and s o m e t i m e s it was as low as 10%. Similar results h a v e been reported by O r e m l a n d et al. [22] in intertidal sediments and b y C a p o n e a n d T a y l o r [7] in tropical seagrass sediments. This m a y be due to i n c o m p l e t e n e s s o f the CEH 2 block or to other processes using nitrate, or both. One possibility is the dissimilatorY reduction o f nitrate to a m m o n i u m . In m a r i n e sediments, the p r o p o r t i o n of nitrate being reduced to a m m o n i u m rather than to N2 has been s h o w n to be as m u c h as 70% o f the total [18, 30], a n d to increase at low redox potential [6] and at low nitrate concentration [17]. In the saltmarsh sediment, 89% o f the a d d e d nitrate was accounted for in one experiment, 51% going to a m m o n i u m and 38% to N20. T h e l l % u n a c c o u n t e d for m a y simply h a v e been due to e x p e r i m e n t a l error, to the fact that we m e a s u r e d only free a m m o n i u m , to other processes using nitrate (e.g., assimilation), or to incompleteness o f the C2H2 block. T h o u g h the C2H2 block m e t h o d has been used with a p p a r e n t success by Yoshinari et al. in soils [40] and by Sorensen in m a r i n e s e d i m e n t s [29], and
Nitrate for Acetylene Inhibition of NzO
151
4o
ao,
z
1
0
0 10
I
I
~
I
1
12
I
_t
14
I
1
16
1
t
18
,
9
20
22
120 - 100
~ 8C
!oo
!
20 ~
0
.a-
2
9
I.
i
4
0
I
L ,
I
2 4 TIME (h)
0
2
4
6
Fig. S. Effectsof preincubation of saltmarsh sediment; 200 #M nitrate added at zero time. (A) Preincubationwith (11)and without ({3)acetylene. Panels 8, C, and D show nitrous oxide reduction In reincubations taken from the parent preincubations at the points marked in A: preincubation and reincubation with acetylene (ml);preincubation without and reincubation with acetylene (0); l~reincubationwith and reincubation without acetylene ([3);preincubation and reincubation without acetylene (O). Bars represent standard error of the mean.
has been validated by c o m p a r i s o n with 13N in soils [28] and with ~SN in sediments [27, 29], some inconsistencies have been observed [23]. Y e o m a n s and Beauchamp [38], working with soils, noted that N 2 0 began to disappear from incubations with 1 and 10% C2H2 after 100 and 160 hours, respectively. Van Raalte and Patriqu~n [361 obtained inconsistent results with saltmarsh sediment. In one experiment, accumulated N 2 0 disappeared between 4 and 8 Clays in the presence o f 10% C2Hz; in a n o t h e r experiment, the N 2 0 persisted. Our own previous investigations [26] indicated that in Flax P o n d saltmarsh Sediment, N~O accumulated only temporarily in the presence o f C2H2; time
152
J.M. Slater and D. G. Capone
A
240
~0
200 160
20
120 '3
30
0,~,~ 10 !
40 o
W
