Environ. Sci. Technol. 2005, 39, 7134-7140
Biogeochemical Controls on Diel Cycling of Stable Isotopes of Dissolved O2 and Dissolved Inorganic Carbon in the Big Hole River, Montana S T E P H E N R . P A R K E R , * ,† SIMON R. POULSON,‡ CHRISTOPHER H. GAMMONS,§ AND MICHAEL D. DEGRANDPRE| Department of Chemistry and Geochemistry, Montana Tech of The University of Montana, Butte, Montana, Department of Geological Sciences, University of NevadasReno, Reno, Nevada, Department of Geological Engineering, Montana Tech of The University of Montana, Butte, Montana, and Department of Chemistry, The University of Montana, Missoula, Montana
Rivers with high biological productivity typically show substantial increases in pH and dissolved oxygen (DO) concentration during the day and decreases at night, in response to changes in the relative rates of aquatic photosynthesis and respiration. These changes, coupled with temperature variations, may impart diel (24-h) fluctuations in the concentration of trace metals, nutrients, and other chemical species. A better understanding of diel processes in rivers is needed and will lead to improved methods of data collection for both monitoring and research purposes. Previous studies have used stable isotopes of dissolved oxygen (DO) and dissolved inorganic carbon (DIC) as tracers of geochemical and biological processes in streams, lakes, and marine systems. Although seasonal variation in δ18O of DO in rivers and lakes has been documented, no study has investigated diel changes in this parameter. Here, we demonstrate large (up to 13‰) cycles in δ18O-DO for two late summer sampling periods in the Big Hole River of southwest Montana and illustrate that these changes are correlated to variations in the DO concentration, the C-isotopic composition of DIC, and the primary productivity of the system. The magnitude of the diel cycle in δ18ODO was greater in August versus September because of the longer photoperiod and warmer water temperatures. This study provides another biogeochemical tool for investigating the O2 and C budgets in rivers and may also be applicable to lake and groundwater systems.
Introduction Many rivers and streams undergo diel (24-h) concentration fluctuations of metals and other chemical species (1-9). Cd, * Corresponding author phone: (406)496-4185; fax: (406)496-4135; e-mail:
[email protected]. † Department of Chemistry and Geochemistry, Montana Tech of The University of Montana. ‡ University of Nevada-Reno. § Department of Geological Engineering, Montana Tech of the University of Montana. | The University of Montana. 7134
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 18, 2005
Mn, and Ni were shown to undergo diel concentration cycles of up to 119, 306, and 167%, respectively, in alkaline streams in Idaho and Montana (6). Dissolved zinc concentrations in Prickly Pear Creek, MT, underwent a 500% change in one 24-h sampling period (9). These diel processes have been shown to occur in all seasons and in many cases the 24-h changes are larger than the seasonal changes in physical and chemical parameters (9). Previous studies have shown that the δ18O of dissolved oxygen (DO) in an aquatic system can vary in relation to the degree of oxygen saturation (10-12). Photosynthesis, respiration, and air-water gas exchange have a direct effect on both the concentration and isotopic composition of DO. In general, photosynthesis by aquatic plants produces DO that is similar in isotopic composition to the source water (13, 14). Since the δ18O of most natural waters is lower than the δ18O of air, photosynthesis tends to lower the δ18O of DO in a river. In contrast, community respiration causes a kinetic isotopic fractionation of molecular oxygen that enriches DO in the water column with the heavier isotope (13, 15). Equilibrium gas exchange across the air-water interface results in DO that is about 0.7‰ heavier than atmospheric oxygen (16, 17). Since the global average δ18O of atmospheric oxygen is +23.5‰ (18), this means that DO in isotopic equilibrium with air should have δ18O ) +24.2‰. However, in instances where the DO concentration of water is rapidly changing because of in-stream biological processes, it is likely that δ18O-DO values will deviate substantially from this equilibrium value. Previous studies have used δ18O-DO measurements to quantify the relative rates of photosynthesis and respiration in rivers and lakes. Quay et al. (10) observed δ18O-DO values that ranged from 15 to 30‰ in 23 rivers and lakes in the Amazon Basin, Brazil, and found that these systems were primarily respiration dominated. Under-ice DO measurements from a 1000-km transect of the Athabasca River, Canada, (11) showed an increasing enrichment in δ18O-DO with decreasing DO concentration. Investigations of DO and δ18O-DO in Lake Superior, Canada, (12) found spatial and temporal variations on a seasonal scale and also demonstrated that trophic levels had significant effects on the δ18ODO. Studies in the Ottawa River, Canada, and two smaller watersheds (14) indicated that these river systems were respiration dominated year-round and that temperature had the same relative effect on both photosynthesis and respiration rates. These studies have shown that it is possible to combine the chemical and isotopic mass balance equations for O2 to calculate the respiration-to-photosynthesis ratio (R:P) in aquatic systems and how this ratio changes with the seasons. Previous investigations have made a steady-state approximation, assuming that diel changes in the concentrations of [18O2] and [O2] were negligible (10, 19). We show that highly productive rivers can experience large changes in DO concentration and isotopic composition over a diel period. To our knowledge, there have been no previous investigations of diel changes in δ18O-DO in any aquatic environment other than several preliminary studies (2022). Similarly to DO, the isotopic composition of dissolved inorganic carbon (DIC) in a river is determined by a balance of the effects of photosynthesis, respiration, and gas exchange, as well as carbonate mineral precipitation or dissolution. During photosynthesis, aquatic plants uptake 12CO2 at a faster rate than 13CO2 (23). Conversely, plant respiration and soil microbes typically produce CO2 that has an isotopic signature similar to that of the indigenous vegetation (24). In temperate 10.1021/es0505595 CCC: $30.25
2005 American Chemical Society Published on Web 08/18/2005
FIGURE 1. Site map showing the Dickie Bridge sampling area of the Big Hole River located in southwestern Montana. regions, C3 plants are the dominant species in both aquatic and terrestrial regimes with δ13C values in the range of -20 to -30‰ (24). Thus, biogenically produced CO2 is normally isotopically light relative to atmospheric CO2 which in the western United States is typically in the range of -7 to -8.5‰ (25). Equilibrium fractionation of atmospheric CO2 should produce dissolved carbon dioxide with δ13C ) -8.2 to -9.7‰ and bicarbonate ion with δ13C ) -1 to +3 ‰ (24). In streams with alkaline pH, the DIC is dominated by bicarbonate. Groundwater inputs can also be an important variable in controlling the isotopic composition of DIC and, to a lesser extent, DO in a river. Groundwater will typically contain DIC that is isotopically light and DO that is isotopically heavy, because of the influence of plant and microbial respiration in the subsurface. Diel changes in δ13C-DIC have been previously studied in calcium carbonate depositing springs (26, 27). It was determined in these cases that off-gassing of carbon dioxide and precipitation of calcite were the major processes influencing the inorganic carbon mass balance and carbon isotope composition of the stream. Site Description. This study details two samplings (August 21 and September 24-25, 2004) that were conducted in the Dickie Bridge area of the Big Hole River, about 40 km southwest of Butte, Montana (Figure 1). The Big Hole River is a headwater tributary to the Missouri River. It is undammed and drains a sparsely populated, high elevation basin (∼1900 m above sea level) of approximately 7200 km2. This river is relatively pristine, and there is little historical impact from mining or industrial sources. The principal activities in the basin are farming, ranching, and recreation. Additionally, the Big Hole River is home of the last self-sustaining population of fluvial Arctic Grayling in the lower 48 states, a species which may soon be listed as endangered (28). The river and its tributaries are important sources of water for municipal and agricultural uses. Previous work (29-31) has summarized the general geochemical characteristics of the Big Hole River. The average values for selected physical and chemical parameters measured during previous diel studies at Dickie Bridge are given in Table 1. The results show very large diel ranges in pH (up to 2.0 standard units) and DO (up to 5 mg L-1) during each sampling event. Overall, the Big Hole River at Dickie Bridge can be classified as a Na-Cabicarbonate water, with alkaline pH and low to moderate alkalinity. All of the samples discussed in this paper were collected near Dickie Bridge (N45° 51.86′ W113° 5.16′, 1750 m elev.), in the middle reach of the Big Hole River (Figure 1). This
sampling site was chosen on the basis of the very large diel pH and DO fluctuations documented in previous work (31) as well as on the lack of obvious ground or surface water contributions in the study reach. The first field study reported here took place on August 21, 2004. Thirteen samples were collected hourly from 0700 to 1800 h. The second field study occurred from 1700 h on Sept 24 to 1800 h on Sept 25, 2004. Nineteen samples were collected hourly during the day and at longer intervals (1.5-3 h) during the night. Weather during the August 21 sampling included scattered clouds and a brief thundershower during the afternoon while no clouds were present during the entire September sampling. The large diel fluctuations in pH and dissolved oxygen concentration that have been observed at Dickie Bridge during summer, low-flow conditions (Table 1) are attributed in part to the slow, laminar flow of water through this reach and also to loading of nutrients from agricultural runoff leading to periodic algal blooms and a chronically thick layer of periphyton coating boulders in the streambed. The partial pressure of dissolved CO2 (pCO2) was measured directly using a SAMI-CO2 autonomous sensor (33, 34) on August 10-11, 2004, with an average value of 15.4 µatm (range: 8.6-24.5) (35). Six weeks later, the average pCO2 calculated from the measured pH, temperature, and alkalinity values during the September sampling was 32 µatm. In comparison, the computed partial pressure of CO2 in equilibrium with air at the site elevation is ∼250 µatm. The low pCO2 values demonstrate the high primary productivity of the Big Hole River in late summer and furthermore indicate that the river was highly undersaturated with respect to atmospheric carbon dioxide during all of the field work of this study.
Methods All sampling bottles were acid-washed (5% HNO3) and rinsed with deionized water three times prior to field work. Water samples for δ18O-H2O and δ13C-DIC analysis were collected by hand using a 60-mL HDPE syringe which was rinsed three times with river water before sample collection. Water was sampled from the main stem of the river in a well-mixed, rapidly flowing reach approximately 3 m from shore and at a depth of ∼15 cm (halfway between surface and bottom). The water was immediately filtered on site using 0.2-µm PES syringe filters. Samples for δ18O-H2O determination were collected in 4-mL glass vials with plastic conical liners and were analyzed after Epstein and Mayeda (36) by the CO2 equilibration technique, using a Micromass Aquaprep device interfaced to a dual inlet Micromass Isoprime stable isotope ratio mass spectrometer. Samples for δ13C-DIC were collected by filtering water into 100-mL glass bottles for precipitation of dissolved inorganic carbon as SrCO3 after Usdowski et al. (26) and were analyzed using a Eurovector elemental analyzer interfaced to a Micromass Isoprime stable isotope ratio mass spectrometer after Harris et al. (37). Collection and analysis of samples for determination of δ18O-DO followed the methods of Wassenaar and Koehler (11). Unfiltered water for δ18O-DO was collected in preevacuated 125-mL serum bottles to which 50 µL of saturated HgCl2 was added prior to evacuation. The samples were prepared for δ18O-DO analysis using a headspace equilibration technique (11). At the field site, all sample bottles were stored on ice in sealed plastic bags and were returned to the laboratory for processing immediately following the field work. Isotope values are reported in units of ‰ in the usual δ notation versus VSMOW for oxygen and VPDB for carbon. Replicate analyses indicate an analytical uncertainty of (0.1‰ for δ18O-DO and (0.05‰ for δ13C-DIC. In-situ temperature, pH, dissolved oxygen concentration (mg L-1), and % saturation were measured at each sampling time with an In Situ Troll 9000 data sonde and a WTW 340i hand held meter. The instruments were calibrated according to the VOL. 39, NO. 18, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7135
TABLE 1. Selected Physical and Chemical Parameters for the Big Hole River at Dickie Bridgea date
ref
flow
pH (range)
DO (range)
temp. (range)
8-15 to 8-16, 2000 8-3 to 8-4, 2001 8-6 to 8-7, 2002 8-10 to 8-11, 2004
30 30 30 32
3.1 (est) 5.5 6.1 4.4
8.93 (7.73-9.77) 8.09 (7.16-8.97) 9.50 (8.90-10.0) 9.28 (9.01-9.58)
8.0 (5.6-10.5 ) 7.8 (5.5-10.5 ) 7.4 (5.1-9.8) 7.5 (5.8-10.1 )
14.7 (11.0-18.5) 17.5 (14.7-20.9) 16.8 (14.4-19.2) 18.2 (13.0-23.0)
date
Ca2+
Mg2+
Na+
K+
SO42-
Cl-
HCO3-
CO32-
8-15 to 8-16, 2000 8-3 to 8-4, 2001 8-6 to 8-7, 2002 8-10 to 8-11, 2004
0.36 0.29 0.28 0.36
0.14 0.11 0.11 0.11
0.45 0.45 0.44 0.56
0.090 0.077 0.072 0.095
0.039 0.050 0.033 0.38
0.054 n.a. n.a. 0.059
1.46 0.88 0.75 1.04
0.054 0.010 0.148 0.140
a
All values are 24-h averages. Temp. (°C), DO (mg L-1), data for major ions are in mmol L-1. Flow in m3 s-1. n.a. ) not analyzed.
FIGURE 2. Diel changes in chemical and physical parameters during the August 21, 2004 (a and b) and the September 24-25, 2004 (c and d) samplings. PAR (photosynthetically active radiation, µE m-2 s-1), pH, dissolved oxygen (DO, % saturation), and water temperature (°C) are plotted vs local time (MDT, GMT -0600). The shaded areas represents the nonphotosynthetic period as defined by the PAR. manufacturer’s recommendations using water-saturated air for the DO probe and standard buffers (pH 7 and 10) for the pH probe. Post-deployment calibration checks showed that the Troll 9000 DO probe was within 2% of pre-deployment calibration. A post-deployment calibration check of the pH probe indicated that the reproducibility was within (0.1 pH units. Photosynthetically active radiation (PAR) values were recorded with a Li-COR, LI-192SA detector and data logger using the manufacturer’s calibration. All time is reported as local time (MDT, GMT -0600).
Results Chemical and Physical Parameters. In-situ measurements of temperature, pH, dissolved oxygen concentration, and PAR (µE m-2 s-1) (Figure 2) for the August and September sampling events showed large fluctuations in pH, temperature, and DO. The amplitude of the cycles was somewhat more pronounced in August versus September. During the day, photosynthesis produced O2 and consumed CO2, which caused the pH of the water column to increase. These relationships reversed at night when respiration was the only biological process. The larger amplitude in the diel pH and DO curves in August is explained by the longer period of solar illumination and the warmer water temperatures at this time (the maximum temperature on Aug 21 was 6 °C 7136
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 18, 2005
higher than on Sept 25), which increased the rates of biological reactions. δ18O-DO and δ13C-DIC Cycles. The average δ18O-H2O values of Big Hole River water during the late August and September samplings were -16.1‰ ((0.1, n ) 13) and -17.0‰ ((0.04, n ) 19), respectively. The short-term temporal variation in δ18O-H2O during both sampling events was within the precision of the isotopic analysis. In contrast, large diel swings in δ18O-DO were observed in both August and September (Figure 3), with O2 becoming isotopically lighter during the day and heavier at night. In September, the range from the minimum to maximum δ18O-DO was roughly 7‰ (18.7-26.4‰), whereas the total range in August was g13‰ (12.4-25.5‰). The greater isotopic variation in August versus September is partly explained by higher biological activity in the warmer season as well as by the lower solubility of O2 in warm versus cold water. For both sampling events, the DO saturation level was inversely correlated to δ18O-DO (R2 ) 0.95, p < 0.0001 in August; R2 ) 0.88, p < 0.0001 in September). During the day, the δ18O of dissolved oxygen decreased and was much lower than that of atmospheric O2 in equilibrium with DO (+24.2 ‰), indicating that photosynthesis was producing isotopically light oxygen with δ18O equal to the water value faster than equilibrium could be established via gas exchange across
FIGURE 3. Diel changes in δ18O-DO and δ13C-DIC in the Big Hole River during August and September of 2004. Direction of isotope changes as influenced by photosynthesis and respiration is shown on each respective axis. Approximate δ13C and δ18O values for air and water are shown on the respective axes. the air-water interface. These results presented here show a larger range of δ18O-DO at one location over a 24-h period than reported for 16 rivers in the Amazon Basin (10). The δ18O-DO in that study was 24-30‰, with a mean O2 saturation of 64% ((21%) and a mean δ18O-DO of 26‰, which was significantly above atmospheric saturation, indicating respiration-dominated systems. The smaller range in δ18ODO in the Amazon study could partially be due to the much lighter δ18Ο-Η2Ο of the Big Hole River (-16 to -17‰) versus the Amazon River water (∼ -5‰). In the Athabasca River, O2 saturation ranged from 120 to 45% while the δ18O-DO varied from 18 to 27‰ (11). The 18O signature of the source water in this system was similar to the Big Hole River with a δ18Ο-Η2Ο of -19 to -21‰. The isotopic composition of DIC showed a diel variation of 1.5‰ during the September sampling (Figure 3). The smaller variation in δ13C-DIC relative to δ18O-DO is explained by the approximately 5-fold higher concentration of DIC in the water column versus DO. δ13C-DIC began decreasing at 1800 h (1-2 h after δ18O-DO began to increase) and reached a minimum at 0630 h. This trend is attributed to the production of isotopically light biogenic CO2 by respiration and the absence of any photosynthesis. The δ13C-DIC never exceeded a maximum value of about -10‰ during the 24-h sampling period. This indicates that the DIC never approached equilibrium with atmospheric CO2 (∼0‰) and that the contribution of biogenically light carbon by community respiration and/or upstream groundwater inputs had a large influence on the dissolved inorganic carbon isotope ratio. Photosynthesis began to affect the system after 0630 h, causing the δ13C-DIC to increase as a result of kinetic fractionation associated with the preferential uptake of 12CO2 by aquatic plants. The 1.5‰ (-10.0 to -11.5‰) change in δ13C-DIC reported in this study is similar to a variation of 1.0‰ (-11.3 to -12.3‰) for the nearby Clark Fork River, MT, during a diel sampling from July 31 to August 1, 2003 (38). Following Quay et al. (10), the rate of change in oxygen concentration can be represented by eq 1:
d[O2] G ) ([O2]s - [O2]w) - R + P + A dt Z
()
(1)
where G is the gas-transfer rate for oxygen (m h-1), Z is the average river depth (m), [O2]s is the dissolved oxygen concentration at saturation (mol L-1), [O2]w is the measured DO of the water (mol L-1), d[O2]/dt is the measured hourly change in DO (mol L-1 h-1), R is the change in DO because of respiration (mol L-1 h-1), P is the photosynthetic production of oxygen (mol L-1 h-1), and A is the accrual of O2 from
FIGURE 4. Plot of d[O2]/dt (µmol L-1 h-1) vs ([O2]s - [O2]w) (µmol L-1) during nighttime for August (a) and September (b). Aeration coefficients (G/Z) determined from the slope are shown. ground or surface water sources. In this study, inputs of O2 from groundwater were assumed to be negligible, on the basis of the absence of any obvious springs and a constant specific conductance along the shore of the river for 1 km upstream of the study area. Although DO can also be consumed by oxidation of dissolved metals such as Fe2+ or Mn2+, the concentrations of these solutes in the Big Hole River are too low to influence the diel O2 mass balance (29). Numerous gas-transfer rates (G) for oxygen are published in the literature, but these values are highly sensitive to changes in the local river conditions. By rearrangement of eq 1, the ratio of G/Z or the “aeration coefficient” can be determined from the slope of a plot of d[O2]/dt versus ([O2]s - [O2]w) for data collected at night when P is zero. This approach, adapted from Odum (39), was chosen since it uses in-situ data for the calculation. While many semiempirical methods are available for calculating the gas-transfer coefficient, it is typically more reliable to use site-specific data when available (40). Aeration coefficients of 0.51 h-1 and 0.59 h-1 were obtained with this method for August and September, respectively (Figure 4). These values compare to a range of 2.0-0.013 h-1 reported by Bennett and Rathburn (41) for streams with average flow velocities in the range of the Big Hole River (∼0.4 m s-1). The average depth (weighted for flow) was calculated to be 0.35 and 0.40 m for August and September, respectively. On the basis of the aeration coefficient, this yields gas-transfer velocities of 0.18 and 0.24 m h-1 for August and September. These values compare to 0.47 m h-1 calculated for the nearby Clark Fork River near Missoula, MT, (42) which had a flow approximately 6.5 times larger than the Big Hole River. The net productivity (P - R, or Pnet) of the river was determined by rearranging eq 1 to yield eq 2.
(P - R) ) Pnet )
d[O2] - (G/Z)([O2]s - [O2]w) dt
(2)
The Pnet value calculated in this manner is the change in O2 concentration per hour because of biological processes alone (i.e., before gas exchange) and assumes that no O2 accrual VOL. 39, NO. 18, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7137
d[18O2] G R ([O ] ‚18:16Oa‚Rs - [O2]w‚18:16ODO) ) dt Z g 2s R‚18:16ODO‚Rr + P‚18:16Ow‚Rp (3)
( )
FIGURE 5. Diel changes in net productivity (Pnet, O2 µmol L-1 h-1) and δ18O-DO during August (a) and September (b) of 2004. Data for Pnet were corrected for gains or losses of O2 to the atmosphere by gas exchange. is occurring from inputs of other water sources. Figure 5a and 5b shows the net productivity and the δ18O-DO for both samplings. The maximum Pnet occurred at 1300 h (55 µmol O2 L-1 h-1) and at 1530 h (45 µmol O2 L-1 h-1) for August and September, respectively. The minima in the δ18O-DO measurements occurred at 1430 and 1400 h in August and September, respectively, and were approximately in phase with the maxima in Pnet. In September, the average rate of consumption of O2 because of respiration from 2030 h to 0700 h was 5.1 µmol L-1 h-1 ((0.7). During the same time period, the average δ18O-DO was 26.0‰ ((0.4), which is heavier than expected for DO in equilibrium with atmospheric oxygen (24.2‰) because of the influence of community respiration. Integrating the area below the zero line for the net productivity curve yielded a total nighttime O2 loss of 59 µmol L-1 during the 12.5-h period from 2000 to 0800. The total daytime production of O2 was 344 µmol L-1 over the remaining 11.5 h. The net O2 balance due to photosynthesis and respiration for the 24-h period of the September sampling was +285 ((20) µmol L-1 da-1, indicating that photosynthesis was the dominant biological process over the 24-h period with a net contribution of O2 to the atmosphere. The total net daytime productivity during the August sampling was 332 ((20) µmol L-1 da-1. The brief thundershower during the afternoon of August 21 decreased the light intensity available for photosynthesis such that the overall daytime productivity in August was similar to September. On the basis of the average river depths, these net productivity values correspond to O2 fluxes of 116 and 91.2 mmol m-2 da-1 for August and September, respectively. In comparison, primary productivity values reported for streams in Idaho, Michigan, Oregon, and Pennsylvania had a range of fluxes from 3.1 to 200 mmol O2 m-2 da-1 (43). Munn and Brusven (44) measured oxygen fluxes ranging from 25 to 100 mmol O2 m-2 da-1 for an unregulated reach of the North Fork of the Clearwater River, ID, which is a similar mountain river to the Big Hole. The 18O2 concentration and δ18O-DO composition can be related by eq 3 from Quay et al. (10). 7138
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 18, 2005
The term 18:16Oa represents the ratio of 18O to 16O in air; Rg is the ratio of the gas-transfer velocities of 18O-16O versus 16O-16O of 0.9972 (45); R is the ratio of the solubility of 18Os 16O versus 16O-16O in water of 1.0007 at 28 °C (17); 18:16O DO is the ratio of 18O to 16O in the dissolved oxygen determined from the measured values of δ18O-DO; Rr is the ratio of the rate of consumption of 18O-16O versus 16O-16O by respiration of 0.982 ( 0.003 (10); 18:16Ow is the ratio of 18O to 16O in the river water determined from the measured δ18O-H2O; and Rp is the ratio of the rate of consumption of H218O versus H216O by photosynthesis of 1.000 ( 0.003 (13). The 18:16Oa was calculated from two measurements of δ18O-air during September with an average value of 23.48‰. The absolute ratio of 18:16O in VSMOW used was 0.002052 (46). The aeration coefficient (G/Z), respiration (R), and primary productivity (P) have been discussed previously. The rate of change in the concentration of 18O-16O (d[18O2]/dt) was modeled for September using eq 3. The value of R used in this calculation was 5.1 µmol L-1 h-1, the average nighttime respiration rate determined previously. The daytime productivity rate (P) was determined from Pnet assuming R is constant. This approximation makes the assumption that photorespiration during the day is negligible. Figure 6 shows the modeled rate using eq 3 for d[18O2]/dt versus the rate determined from the measured [O2] and δ18O-DO. The modeled rate is well correlated (R2 ) 0.88, p < 0.0001) with the rate determined from field data. The calculated rate reaches a maximum value at 1000 h and the rate based on field data peaks at 1100 h. This may be due to photorespiration which can retard the photosynthetic rate of O2 production as it approaches a daytime maximum. The fractionation factors were used as cited above. Although isotope fractionation (R) changes with temperature, it has been assumed that these changes are negligible for the small temperature ranges of interest. The value for Rr referenced above and used in this calculation was measured for the Amazon Basin. Other published values of Rr for microalgae and bacterial respiration range from 0.972 to 1.000 (47, 48). Using Rr values across this range did not change the modeled result. Although the diel changes in δ18O-DO and δ13C-DIC reported here are clearly related to the daily cycle of photosynthesis and respiration, on the basis of our results, the timing of the maxima and minima of these two isotopic systems are not in phase. Thus, when plotted against each other, the data from the September sampling form a closed loop instead of a single reversible linear or curved path (Figure 7). The circular pattern is a result of the difference in timing and rates of the O2 and CO2 fluxes in to and out of the water column. For example, the trend in isotope composition from 1400 h to 1830 h can be explained by the fact that the water is supersaturated in O2 and will be driven toward equilibrium with heavier atmospheric oxygen. The rate of change in 18O2 reached a maximum at 1100 h (Figure 6) suggesting that after this time gas exchange is comparable to DO production. Additionally, the net productivity began decreasing after 1530 h (Figure 5) indicating that photosynthesis was less dominant after this time. During the same time period, CO2 was undersaturated (see Site Description), and both gas exchange and photosynthesis will have resulted in isotopic enrichment of DIC. From 1830 to 2030 h, respiration contributed to an increase in δ18O-DO and a decrease in δ13C-DIC. After 2030 h, the rate of DO consumption by respiration and gas exchange became approximately balanced such that the δ18ODO was relatively constant overnight. The δ13C-DIC continued to decrease during the night because gas exchange does not
FIGURE 6. The change in 18O2 concentration with time (d[18O2]/dt, µmol L-1 h-1) during September calculated from eq 3 versus the rate determined from field data.
the factors controlling the rate of change of 18O to photosynthesis, respiration, and gas exchange. This study has important implications for future studies of the biogeochemistry of rivers. Our results clearly show that the concentration and isotopic composition of dissolved O2 and dissolved inorganic carbon are dynamic parameters that can vary significantly over short time periods. Thus, for highly productive rivers and lakes, the assumption of chemical and isotopic steady state is unlikely to be valid, making estimates of net productivity on the basis of a single instantaneous sample unreliable. Furthermore, collection of δ18O-DO in combination with δ13C-DIC has the potential to lend new insight into the timing and magnitude of mechanisms that influence the biogeochemistry of rivers. Further field studies are planned to examine seasonal changes in the timing and magnitude of diel isotopic fluctuations in δ18O-DO and δ13C-DIC of the Big Hole River and to link these changes to diel variations in water chemistry, including instream cycling of nutrients and trace metals. Additionally, it should be possible to extend the methods employed here to other terrestrial aquatic systems, including lakes and wetlands, as well as confined and unconfined groundwater aquifers.
Acknowledgments We thank B. Parker for field and technical assistance. Funding for S. Parker and C.H.G. was provided by EPA-EPSCoR and the Montana Board of Research and Commercialization. S. Poulson was funded in part by the NSF. M.D.D was funded by NSF EAR-0337460.
Literature Cited FIGURE 7. Plot of δ18O-DO vs δ13C-DIC over the 24-h period in September. The starting and ending points are shown as well as the time during the field experiment. Arrows illustrate the direction of time during the sampling. counterbalance respiration because of the comparatively slow CO2 exchange versus O2 (Figure 3). Starting at ∼0800 h, photosynthesis becomes the dominant process and δ18ODO decreased while δ13C-DIC increased. During the time periods 1830-2030 and 0800-1400, both δ18O-DO and δ13CDIC show a reasonable negative correlation (R2 ) 0.83 and 0.92, respectively) during the time when gas exchange was less important since DO was closer to saturation. In summary, this study has documented substantial and reproducible diel changes in δ18O-DO in two field studies on the Big Hole River, MT, as well as a significant diel change in δ13C-DIC in one of these studies. The δ18O-DO cycle was inversely related to changes in the percent saturation of the dissolved oxygen and δ13C-DIC of the system. A significant result is the fact that the decrease of δ18O-DO during the day, to a value well below atmospheric equilibrium, was due to the production of DO by photosynthesis from substrate water that had a much lower δ18O than atmospheric O2. The increase at night was in response to preferential consumption of light O2 by respiring organisms coupled with inward diffusion of isotopically heavy O2 from the atmosphere. The nighttime δ18O-DO stabilized above the atmospheric equilibrium value because of the effect of community respiration balanced by gas exchange. The changes in carbon and oxygen isotope ratios were fundamentally controlled by the net productivity (balance of photosynthesis and respiration) of the aquatic ecosystem. Using oxygen isotope fractionation factors from the literature, satisfactory agreement was demonstrated between the predicted versus modeled rate of change in 18O2 of the water column. This result is important since the model constrains
(1) McKnight, D. M.; Kimball, B. A.; Bencala, K. E. Iron photoreduction and oxidation in an acidic mountain stream. Science 1988, 240, 637-640. (2) Fuller, C. C.; Davis, J. A. Influence of coupling of sorption and photosynthetic processes on trace element cycles in natural waters. Nature 1989, 340, 52-54. (3) Bourg, A. C. M.; Bertin, C. Diurnal variations in the water chemistry of a river contaminated by heavy metals. Natural biological cycling and anthropic influence. Water, Air, Soil, Pollut. 1996, 86, 101-116. (4) Brick, C. M.; Moore, J. N. Diel variation of trace metals in the upper Clark Fork River, Montana. Environ. Sci. Technol. 1996, 30, 1953-1960. (5) Sullivan, A. B.; Drever, J. I.; McKnight, D. M. Diel variation in element concentrations, Peru Creek, Summit County, Colorado. J. Geochem. Explor. 1998, 64, 141-145. (6) Nimick, D. A.; Gammons, C. H.; Cleasby, T. E.; Madison, J. P.; Skaar, D.; Brick, C. M. Diel cycles in dissolved metal concentrations in streams s Occurrence and possible causes. Water Resour. Res. 2003, 39, 1247, doi: 10.1029/WR001571. (7) Jones, C. A.; Nimick, D. A.; McCleskey, B. Relative effect of temperature and pH on diel cycling of dissolved trace elements in Prickly Pear Creek, Montana. Water, Air, Soil, Pollut. 2004, 153, 95-113. (8) Gammons, C. H.; Nimick, D. A.; Parker, S. R.; Cleasby, T. E.; McClesky, R. B. Diel behavior of iron and other heavy metals in a mountain stream with acidic to neutral pH: Fisher Creek, MT, USA. Geochim. Cosmochim. Acta 2005, 69, 2505-2516. (9) Nimick, D. A.; Cleasby, T. E.; McClesky, R. B. Seasonality of diel cycles of dissolved trace metal concentrations in a Rocky Mountain Stream. Environ. Geol. 2005, 47, 603-614. (10) Quay, P. D.; Wilbur, D. O.; Richey, J. E.; Devol, A. H. The 18O:16O of dissolved oxygen in rivers and lakes in the Amazon Basin: Determining the ratio of respiration to photosynthesis rates in freshwaters. Limnol. Oceanogr. 1995, 40, 718-729. (11) Wassenaar, L. I.; Koehler, G. An on-line technique for the determination of the delta O-18 and delta O-17 of gaseous and dissolved oxygen. Anal. Chem. 1999, 71, 4965-4968. (12) Russ, M. E.; Ostrom, N. E.; Gandhi, H.; Ostrom, P. H. Temporal and spatial variations in R:P ratios in Lake Superior, an oligotrophic freshwater environment. J. Geophys. Res. 2004, 109, C10S12, doi: 10.1029/2003JC001890. VOL. 39, NO. 18, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7139
(13) Guy, R. D.; Fogel, M. L.; Berry, J. A. Photosynthetic Fractionation of the Stable Isotopes of Oxygen and Carbon. Plant Physiol. 1993, 101, 37-47. (14) Telmer, K.; Veizer, J. Isotopic constraints on the transpiration, evaporation, energy and GPP budgets of a large boreal watershed: Ottawa River basin, Canada. Global Biogeochem. Cycles 2000, 14, 149-166. (15) Kroopnik, P. M. Respiration, photosynthesis and oxygen isotope fractionation in oceanic surface water. Limnol. Oceanogr. 1975, 20, 988-992. (16) Friedman, I.; O’Neil, J. R. Compilation of Stable Isotope Fractionation Factors of Geochemical Interest, 6th ed.; U.S. Geological Survey Proffessional Paper 440-KK; U.S. Geological Survey: Reston, VA, 1977. (17) Benson, B. B.; Krause, D. The concentration and isotopic fractionation of gases dissolved in freshwater in equilibrium with the atmosphere. 1. Oxygen. Limnol. Oceanogr. 1980, 25, 662-671. (18) Kroopnik, P. M.; Craig, H. C. Atmospheric oxygen: Isotopic composition and solubility fractionation. Science 1972, 175, 5455. (19) Wang, X.; Vezier, J. Respiration-photosynthesis balance of terrestrial aquatic ecosystems, Ottawa area, Canada. Geochim. Cosmochim. Acta 2000, 64, 3775-3786. (20) Jamieson, T. S.; Schiff, S. L.; Taylor, W. D. Implementation of a New Isotopic Tool (18O/16O in dissolved O2) for the Management of Oxygen in the Grand River, an Impacted Watershed in S. Ontario. EOS Trans., AGU, 2004, 85 (17), Abstract B41B-04. (21) Silva, S. R.; Doctor, D. H.; Kendall, C.; Chang, C. C.; Fleenor, W. E. Assessing Spatial and Temporal Variability in Aquatic Photosynthesis and Respiration with Oxygen and Carbon Stable Isotopes. EOS Trans., AGU, 2004, 85 (47), Abstract B21C-0904. (22) Tobias, C. R.; Bohlke, J. Diel Variation of Oxygen and Inorganic Carbon in a High Productivity, High Alkalinity Stream: Biological and Geochemical Controls. EOS Trans., AGU, 2004, 85 (47), Abstract H51G-07. (23) Falkowski, P. G.; Raven, J. A. Aquatic Photosynthesis; Blackwell Sciences: Malden, MA, 1997. (24) Clark, I. D.; Fritz, P. Environmental Isotopes in Hydrogeology; Lewis Pub: New York, 1997. (25) NOAA: Climate monitoring and diagnostics laboratory, Carbon cycle greenhouse gases 2004. http://www.cmdl.noaa.gov/ccgg/ iadv (accessed Nov 2004). (26) Usdowski, E.; Hoefs, J.; Menschel, G. Relationship between 13C and 18O fractionation and changes in major element composition in a recent calcite-depositing spring - A model of chemical variations with inorganic CaCO3 precipitation. Earth Planet. Sci. Lett. 1979, 42, 267-276. (27) Spiro, B.; Pentecost, A. One day in the life of a stream - A diurnal inorganic carbon mass balance for a travertinedepositing stream (Waterfall Beck, Yorkshire). Geomicrobiol. J. 1991, 9, 1-11. (28) Montana Department of Fish, Wildlife and Parks: 2004. http:// www.fisheries.org/AFSmontana/SSCpages/grayling_status.htm (accessed Feb 2005). (29) Ridenour, R. A seasonal and spatial chemical study of the Big Hole River, Southwest Montana. M.S. Thesis, Montana Tech of the University of Montana, Butte, MT, 2002. (30) Wenz, A. C. Diurnal Variations in the Water Quality of the Big Hole River. M.S. Thesis, Montana Tech of the University of Montana, Butte, MT, 2003.
7140
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 18, 2005
(31) Gammons, C. H.; Ridenour, R.; Wenz A. Diurnal and longitudinal variations in water quality along the Big Hole River and tributaries during the drought of August, 2000. Mont. Bur. Mines Geol., Open-File Report 424, 39, 2001. (32) Parker, S. R.; Gammons, C. H. Montana Tech of The University of Montana. Unpublished work, Aug 10-11, 2004. (33) Baehr, M. M.; DeGrandpre, M. D. In situ pCO2 and O2 measurements in a lake during turnover and stratification: Observations and modeling. Limnol. Oceanogr. 2004, 49, 330340. (34) DeGrandpre, M. D.; Hammar, T. R.; Smith, S. P.; Sayles, F. L. A submersible autonomous sensor for spectrophotometric pH measurements of natural waters. Limnol. Oceanogr. 1995, 40, 969-975. (35) DeGrandpre, M. D. The University of Montana. Unpublished work, Aug. 10-11, 2004. (36) Epstein, S.; Mayeda, T., Variation of O18 content of waters from natural sources. Geochim. Cosmochim. Acta 1953, 4, 213-224. (37) Harris, D.; Porter, L. K.; Paul, E. A. Continuous flow isotope ratio mass spectrometry of carbon dioxide trapped as strontium carbonate. Commun. Soil Sci. Plant Anal. 1997, 28, 747-757. (38) Parker, S. R.; Gammons, C. H.; DeGrandpre, M. D.; Poulson, S. R. Diel changes in stable carbon isotope ratios and trace element concentrations in the Clark Fork River, MT. EOS Trans., AGU, 2004, 85 (47), Abstract H41F-0344. (39) Odum, H. T. Primary production in flowing waters. Limnol. Oceanogr. 1956, 1, 102-117. (40) Melching, C. S.; Flores, H. E. Reaeration equations derived from USGS database. J. Environ. Eng. 1999, 125 (5), 407-414. (41) Bennett, J. P.; Rathburn, R. E. Reaeration in Open-Channel Flow; U.S. Geological Survey Professional Paper 737; U.S. Geological Survey: Washington, DC, 1975; pp 1-75.. (42) Reynolds, J. C. Distributed in situ gas measurements for the analysis and modeling of biogeochemical changes in the Clark Fork River. M.S. Thesis, University of Montana, Missoula, MT, 2001. (43) Bott, L.; Brock, J. T.; Dunn, C. S.; Naiman, R. J.; Ovink, R. W.; Petersen, R. C. Benthic community metabolism in four temperate stream systems: An inter-biome comparison and evaluation of the river continuum concept. Hydrobiologia 1985, 123, 3-45. (44) Munn, M. D.; Brusven, M. A. The influence of Dworshak Dam on epilithic community metabolism in the Clearwater River, USA. Hydrobiologia 2004, 513, 121-127. (45) Knox, M. P.; Quay, P. D.; Wilbur, D. O. Kinetic fractionation during air-water gas transfer of O2, N2, CH4 and H2. J. Geophys. Res., Oceans 1992, 97, 335-343. (46) Baertschi, P. Absolute 18O content of standard mean ocean water. Earth Planet. Sci. Lett. 1976, 31, 341-344. (47) Kiddon, J.; Bender, M. L.; Orchardo, J.; Caron, D. A.; Goldman, J. C.; Dennett, M. Isotopic fractionation of oxygen by respiring marine organisms. Global Biogeochem. Cycles 1993, 7 (3), 679694. (48) Luz, B.; Barkan, E.; Sagi, Y. Evaluation of respiratory mechanisms with oxygen isotopes: A case study in Lake Kinneret. Limnol. Oceanogr. 2002, 47 (1), 33-42.
Received for review March 21, 2005. Revised manuscript received June 30, 2005. Accepted July 7, 2005. ES0505595
