PUBLICATIONS Journal of Geophysical Research: Biogeosciences RESEARCH ARTICLE 10.1002/2015JG002941 Key Points: • Epilimnion N2O in reservoirs is positively related to nitrite and nitrate • Nitrate predicts whether hypoxic hypolimnia are a net N2O source or sink • Mixing events can lead to pulses of N2O production in reservoirs
Supporting Information: • Supporting Information S1 • Data Set S1 Correspondence to: J. J. Beaulieu,
[email protected]
Citation: Beaulieu, J. J., C. T. Nietch, and J. L. Young (2015), Controls on nitrous oxide production and consumption in reservoirs of the Ohio River Basin, J. Geophys. Res. Biogeosci., 120, 1995–2010, doi:10.1002/2015JG002941. Received 30 JAN 2015 Accepted 30 AUG 2015 Accepted article online 4 SEP 2015 Published online 21 OCT 2015
Controls on nitrous oxide production and consumption in reservoirs of the Ohio River Basin Jake J. Beaulieu1, Christopher T. Nietch1, and Jade L. Young2 1
National Risk Management Research Laboratory, U.S. Environmental Protection Agency, Office of Research and Development, Cincinnati, Ohio, USA, 2Louisville District Water Quality, U.S. Army Corps of Engineers, Louisville, Kentucky, USA
Abstract Aquatic ecosystems are a globally significant source of nitrous oxide (N2O), a potent greenhouse gas, but estimates are largely based on studies conducted in streams and rivers with relatively less known about N2O dynamics in reservoirs. Due to long water residence times and high nitrogen (N) loading rates, reservoirs support substantial N processing and therefore may be particularly important sites of N2O production. Predicting N2O emissions from reservoirs is difficult due to complex interactions between microbial N processing in the oxygen-poor hypolimnion and oxygen-rich epilimnion. Here we present the results of a survey of N2O depth profiles in 20 reservoirs draining a broad range of land use conditions in four states in the U.S. Nitrous oxide was supersaturated in the epilimnion of 80% of the reservoirs and was undersaturated in only one, indicating that reservoirs in this region are generally a source of N2O to the atmosphere. Nitrous oxide was undersaturated in the hypolimnion of 10 reservoirs, supersaturated in 9, and transitioned from supersaturation to undersaturation in 1 reservoir that was monitored periodically from midsummer to fall. All reservoirs with a mean hypolimnion nitrate concentration less than 50 μg N L1 showed evidence of net N2O consumption in the hypolimnion. All reservoirs sampled during lake turnover supported N2O production throughout the water column. These results indicate that N2O dynamics in reservoirs differ widely both among systems and through time but can be predicted based on N and oxygen availability and degree of thermal stratification. 1. Introduction Nitrous oxide (N2O) is a potent greenhouse gas [Myhre et al., 2013] that also contributes to the depletion of ozone in the upper stratosphere [Ravishankara et al., 2009]. In 2011, atmospheric N2O was 19% above its 1750 level, and anthropogenic emissions constitute 35% to 45% of the global total [Myhre et al., 2013], primarily from the conversion of agricultural nitrogen (N) to N2O in soils and waters via microbially mediated N transformations. Denitrification and nitrification have been identified as the major N2O-producing pathways in soils and waters [Firestone and Davidson, 1989; Thuss et al., 2014]. Denitrification is a form of anaerobic respiration in which nitrate (NO3) is sequentially reduced to nitrite (NO2), nitric oxide (NO), nitrous oxide (N2O), and dinitrogen (N2). Nitrous oxide can be released to the environment when the rate of N2O formation exceeds that of N2O reduction to N2. Conversely, denitrification can be a net N2O sink when the rate of N2O reduction exceeds that of N2O production. Controls on denitrification rates include oxygen levels, N and carbon (C) availability, temperature, and pH [Knowles, 1982]. Nitrification is an aerobic, chemolithoautotrophic process in which ammonium (NH4+) is sequentially oxidized to NO2 and NO3. Hydroxylamine, which is produced as an intermediate product during nitrification, can decompose to form N2O which can be released to the environment. Controls on nitrification rates include NH4+, organic C, pH, and temperature [Bianchi et al., 1999; Strauss et al., 2004].
©2015. The Authors. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
BEAULIEU ET AL.
Nitrous oxide emissions from soils have been intensively studied for decades [Stehfest and Bouwman, 2006], but emissions from aquatic ecosystems have received much less attention. Recent estimates indicate that river networks may be the source of ~10% of global anthropogenic N2O emissions [Beaulieu et al., 2011], but most investigations of aquatic N2O emissions have focused on streams [Baulch et al., 2011; Beaulieu et al., 2009; Beaulieu et al., 2008] and rivers [Beaulieu et al., 2010; Rosamond et al., 2011; Venkiteswaran et al., 2014] with relatively less known about N2O emissions from lentic ecosystems (lakes and reservoirs) [Guérin et al., 2008; Hendzel et al., 2005; Huttunen et al., 2002]. Due to their long water residence time, however, lentic
N2O IN RESERVOIRS
1995
Journal of Geophysical Research: Biogeosciences
10.1002/2015JG002941
ecosystems have been long recognized as systems where extensive N processing can occur [Wetzel, 2001], suggesting they may be important N2O sources. Among lentic ecosystems, reservoirs may support particularly high N2O emissions due to higher N loading and processing rates than lakes [Harrison et al., 2009]. Studies from lotic ecosystems indicate that N2O emissions tend to increase with N loading, but it is unclear whether this pattern holds for lentic systems. Unlike lotic ecosystems, which tend to be oxygenated and well mixed, reservoirs can become thermally stratified leading to the development of well-oxygenated shallow waters (epilimnion) and hypoxic deep waters (hypolimnion) which support different types of microbial N transformations. Nitrous oxide biogeochemistry in oxygen-deprived hypolimnetic waters is primarily controlled by denitrification. The factors that determine whether denitrification is a net source or sink for N2O in the hypolimnion are unclear, though nitrate availability appears to be important [Beaulieu et al., 2014b]. Nitrification is the major N2O-producing pathway in oxic epilimnion, but NH4+ availability (the nitrification substrate) can be very low, even in systems with high N loading, due to NH4+ assimilation by phytoplankton [Smith et al., 2014; Tõnno et al., 2005]. The area of rapidly decreasing dissolved oxygen concentration (i.e., oxycline) located at the interface between the hypoxic hypolimnion and oxic epilimnion is yet another distinct biogeochemical zone where the products of aerobic and anaerobic metabolic processes mix, allowing for high rates of N2O production. For example, ammonium (NH4+) produced from organic matter decomposition in the hypolimnion can be rapidly nitrified at the hypoxic-oxic interface. Similarly, NO3 produced via nitrification in the epilimnion can be subsequently denitrified at the hypoxic-oxic interface through coupled nitrification-denitrification. Within the transitional oxycline, nitrification and denitrification are occurring at the limits of their oxygen tolerance which can lead to a large fraction of denitrified/nitrified N being converted to N2O [Firestone et al., 1980; Goreau et al., 1980], further enhancing N2O production rates. Several studies have reported middepth N2O maxima near the oxycline in stratified freshwaters [Deemer et al., 2011; Knowles et al., 1981; Mengis et al., 1997; Yoh et al., 1988], but these middepth N2O maxima are often transient [Beaulieu et al., 2014b] or fail to develop altogether [Yoh et al., 1988]. While previous research has provided considerable insight into N2O biogeochemistry, a broad synthesis of the factors controlling N2O distributions is lacking, in part because of the lack of comparable data collected across systems that represent a range of morphologies and land use conditions. Here we present the results of a cross-site survey of N2O depth profiles in midlatitude/low-elevation reservoirs draining land from four states in the U.S. within the Ohio River Basin. Sixteen of the reservoirs were sampled once during the period of thermal stratification, and three were sampled while the epilimnion and hypolimnion were mixing. One reservoir was sampled several times during the period of thermal stratification and once while the epilimnion and hypolimnion were mixing. The study objective was to use the range of physiochemical characteristics among the different reservoirs to identify key factors controlling the distribution of N2O in reservoir waters. We hypothesized that in the epilimnion, N2O would be supersaturated due to persistent production via nitrification and negligible consumption via denitrification. The degree of N2O saturation would be determined by the balance between N2O production and evasion to the atmosphere. Below the thermocline in hypoxic hypolimnetic waters, we hypothesized that denitrification would be a net N2O source when NO3 concentrations are high, leading to N2O supersaturation, and a net N2O sink when NO3 concentrations are low, leading to N2O undersaturation. We hypothesized that middepth N2O maxima would be associated with the oxycline and would be more likely to occur when thermal stratification was relatively weak, allowing for greater exchange of solutes across the hypoxic-oxic boundary. During the annual erosion of stratification that occurs in the fall, we hypothesized that the mixing of NH4+-rich hypolimnetic waters with oxygenated surface waters would result in a pulse of nitrification and elevated dissolved N2O concentrations.
2. Methods 2.1. Study Sites We sampled 20 reservoirs in the Ohio River Basin draining portions of Indiana, Ohio, Kentucky, and Tennessee (Figure 1). The reservoirs are operated by the United States Army Corps of Engineers (USACE) and are used for numerous purposes including drinking water supply, flood control, water quality, and recreation. The surface areas and storage volumes of the reservoirs ranged from 1–43 km2 and 0.4–42.6 × 106 m3, respectively
BEAULIEU ET AL.
N2O IN RESERVOIRS
1996
Journal of Geophysical Research: Biogeosciences
10.1002/2015JG002941
Figure 1. Location of the reservoirs sampled in this study. Reservoirs and watersheds are represented with black and grey fills, respectively.
(Table 1). The study sites include reservoirs that drain pasture lands (e.g., SRR, 80% pasture), croplands (e.g., 80% cropland in HTR), and forested land (e.g., 84% forest in BHR) and one reservoir drains a developed watershed (i.e., WFR, 74% urban). More information about the reservoirs, including operation plans and historic water levels can be found at the USACE website (http://www.lrl.usace.army.mil/Missions/CivilWorks/WaterInformation.aspx). 2.2. Field Sampling We sampled 19 reservoirs once during the summer and early fall of 2013 (see Table 1 for sample dates). One reservoir, East Fork Lake (EFR; also known as William H. Harsha Lake), was sampled on four dates between 24 July 2013 and 29 October 2013. All lakes were sampled at their deepest point. Water temperature and dissolved oxygen (DO) were measured at 0.3 m intervals near the thermocline, defined as the plane of maximum rate of temperature decrease [Wetzel, 2001], and at 1.5 m intervals throughout the rest of the water column using a data sonde with optical dissolved oxygen sensor (YSI 6920 V2, Yellow Springs, OH, USA). Water samples were collected from 20 depths ranging from 0.2 m above the lakebed to 0.1 m below the water surface using a horizontal Van Dorn sampler (Wildlife Supply Company, Yulee, FL, USA). Atmospheric pressure was measured using a barometer (YSI MDS 650, Yellow Springs, OH, USA). A water sample for dissolved gas analysis was collected from each sampling depth by transferring water from the Van Dorn sampler into 140 mL glass serum vials using a length of tubing. The serum bottles were filled from the bottom, allowed to overflow by 3 times their volume, preserved with 100 μL of a saturated mercury chloride solution, and sealed with a grey butyl septa. Samples were stored at 5°C until analysis. A water sample for nutrient analysis was collected from each depth by withdrawing 30 mL of water from the Van Dorn sampler using a 60 mL syringe. The water was field filtered (0.45 μm pore size) into an acid washed 30 mL high-density polyethylene bottle, stored on ice and analyzed within 24 h, or frozen and analyzed within 7 days. 2.3. Sample Processing and Analysis Nitrous oxide was extracted from the water samples using headspace equilibration [Ioffe and Vitenberg, 1984]. Forty milliliters of ultrahigh purity helium was transferred into the serum bottle, while an equivalent
BEAULIEU ET AL.
N2O IN RESERVOIRS
1997
BEAULIEU ET AL.
37°20′19.11″N 36°53′31.26″N 39°26′34.22″N 39°56′54.74″N 39°29′8.77″N 37°13′44.00″N 39°43′10.04″N 39°29′13.96″N 38°7′5.96″N 39°1′17.40″N
37°15′3.00″N 40°50′47.53″N 39°0′28.81″N 40°42′51.42″N 37°16′45.00″N 38°26′2.39″N 37°37′6.00″N 40°48′25.00″N 38°0′10.20″N 39°15′38.38″N
Green River Lake (GRR) J.E. Roush (HTR) Monroe Lake (MNR) Mississinewa Lake (MSR) Nolin Lake (NRR) Patoka Lake (PRR) Rough River Lake (RRR) Salamonie (SRR) Taylorsville Lake (TAR) West Fork (WFR)
Latitude
Buckhorn Lake (BHR) Barren River Lake (BRR) Brookville Lake (BVR) C.J. Brown (CBR) Caesar Creek (CCK) Carr Creek Lake (CFK) C.M. Harden (CHL) Cagles Mill (CMR) Cave Run Lake (CRR) East Fork Lake (EFR)
Reservoir
N2O IN RESERVOIRS 85°20′17.00″W 85°28′2.89″W 86°30′55.11″W 85°57′21.83″W 86°14′49.00″W 86°42′16.62″W 86°29′59.00″W 85°40′38.00″W 85°18′20.40″W 84°29′50.83″W
83°28′14.58″W 86°7′21.14″W 85°0′0.49″W 83°44′45.18″W 84°3′36.79″W 83°1′57.00″W 87°4′17.79″W 86°54′53.58″W 83°31′49.56″W 84°9′6.89″W
Longitude 2013-08-12 2013-08-06 2013-08-18 2013-08-28 2013-09-25 2013-08-14 2013-09-26 2013-10-08 2013-09-05 2013-07-24, 2013-08-27, 2013-09-24, and 2013-10-29 2013-08-05 2013-09-09 2013-08-20 2013-09-10 2013-08-08 2013-08-22 2013-09-17 2013-10-02 2013-09-16 2013-08-29
Date
33 3 43 13 23 35 20 11 12 1
5 39 21 8 11 2 8 6 31 8
Reservoir Surface 2 Area (km )
Table 1. Reservoir and Watershed Characteristics for the 20 Systems Included in This Study
38.0 1.0 42.6 5.4 14.9 35.8 7.0 2.6 18.8 0.4
1.6 15.4 33.8 7.2 21.9 3.8 3.8 6.3 41.0 19.2
Storage Volume 6 3 (10 m )
23.4 6.6 15.4 18.5 24.0 12.7 18.5 17.6 21.5 3.6
16.7 16.7 31.9 10.3 31.0 20.0 16.4 10.9 21.8 32.8
Max Depth (m)
1797 1778 1166 347 1582 448 1127 1473 939 98
1072 1500 1031 225 643 153 588 796 2186 882
Watershed 2 Area (km )
4.7 80.4 3.5 65.8 13.8 3.7 8.1 0.0 3.9 0.1
0.1 5.6 52.9 62.8 71.3 0.0 72.9 61.5 0.6 53.6
Crops
24.6 3.6 4.2 1.8 29.0 11.5 31.6 80.0 39.7 1.0
0.2 43.9 9.0 16.0 6.6 0.0 5.5 7.8 8.5 9.2
Pasture
3.7 7.3 2.1 9.6 6.5 4.0 4.1 6.9 6.1 74.3
5.3 6.8 12.1 6.6 6.4 6.7 5.8 5.5 4.1 7.0
Urban
60.8 6.6 81.9 14.0 45.6 67.0 50.5 9.2 44.9 0.7
84.3 38.5 22.5 8.7 13.0 76.3 12.9 22.5 78.3 27.8
Forest
Watershed Land Use (%)
Journal of Geophysical Research: Biogeosciences 10.1002/2015JG002941
1998
Journal of Geophysical Research: Biogeosciences
10.1002/2015JG002941
volume of water was allowed to escape the bottle through a vent needle. The bottle was shaken for ≥ 1 h, and 25 mL of headspace gas, at 1 atm, was withdrawn from the serum bottle and transferred to a pre-evacuated 12 mL glass vial (Exetainer, LabCo, UK). Nitrous oxide was measured using a gas chromatograph (Bruker 450, Billerica, MA, USA) equipped with an electron capture detector and autosampler (CTC Analytics, PAL-xt, Zwingen, Switzerland). A 90:10 mixture of argon and methane was used as the carrier gas. The original dissolved N2O concentration was calculated from the measured headspace concentration, the temperature specific Bunsen coefficient, and a mass balance for the headspace equilibration system as described in Beaulieu et al. [2012]. The equilibrium dissolved gas concentration at the time of sampling was calculated from the global average atmospheric N2O concentration (327 ppb) [Tans and Keeling, 2013], the measured barometric pressure, and the water temperature. Dissolved N2O concentration is presented as a saturation ratio, defined as the ratio of measured to equilibrium concentration. Equilibrium N2O concentration ranged from 193–363 μg N2O-N L1, depending on water temperature and atmospheric pressure. Soluble reactive phosphorus (SRP), nitrite (NO2), nitrate (NO3) + NO2, and ammonium (NH4+) were measured with automated colorimetry (Lachat Instruments QuickChem 8000 Flow Injection Autoanalyzer, Loveland, CO, USA) [Sardina, 2000; Smith, 2001; Wendt, 1995]. 2.4. Statistics and Data Analysis We calculated the thermocline strength index (TSI) as an indicator of thermal resistance to mixing of hypolimentic and epilimentic waters near the thermocline [Horne and Goldman, 1994]. TSI ¼ ΔT=Δh
(1)
where ΔT and Δh are the differences in water temperature (°C) and water depth (m), respectively, between the top and bottom of the metalimnion. We used logistic regression to determine if the probability of the occurrence of a middepth N2O maximum was related to TSI. We categorized the epilimnion and hypolimnion of each lake as a site of N2O production, N2O consumption, or negligible N2O processing based on the mean dissolved N2O saturation ratio in each layer. We considered mean saturation ratios greater than 1.2 and less than 0.8 to be indicative of production and consumption, respectively. We assumed that saturation ratios between these values were indicative of negligible N2O processing. In three instances the mean N2O saturation ratio in the hypolimnion exceeded 0.8 due to a few depths with high dissolved N2O concentrations, despite clear evidence of N2O consumption throughout much of the water layer, and these reservoirs were classified as site of hypolimnion N2O consumption. Lakes where the maximum N2O saturation occurred at middepths were also identified (in the sense of Priscu et al. [2008]). Our analysis assumes that N2O saturation ratios > 1.2 or < 0.8 are indicative of biological production or consumption; however, abiotic factors can also cause the saturation ratio to deviate from 1.0, such as the seasonal warming of waters [Venkiteswaran et al., 2014]. For example, water that equilibrated with the atmosphere at 5°C then warmed to 30°C would have a saturation ratio of 2.3 if no air-water gas exchange occurred during the process. Gas exchange in the well-mixed epilimnion is relatively rapid, however. Assuming a conservative gas exchange rate of 2 cm h1 [Cole and Caraco, 1998], the dissolved gas content of a 3 m thick epilimnion would equilibrate with the atmosphere in ~6 days, which is sufficiently rapid to allow excess gas formed by rising water temperatures to off-gas to the atmosphere. While no gas exchange occurs between the hypolimnion and atmosphere, water temperatures in the hypolimnion are relatively stable, often increasing by no more than 3°C during the summer [Beaulieu et al., 2014b], which would increase the saturation ratio by ~ 0.1. We calculated the difference in mean nutrient concentrations between the epilimnion and hypolimnion of all stratified lakes and used a paired t test to determine if the differences were statistically significant. We assessed relationships between dissolved N2O and potential controlling variables using simple linear regression unless the relationship was clearly nonlinear, in which case we modeled the relationship using a two-parameter exponential model: y ¼ að1 ecx Þ
(2)
where y and x are the dependent and independent variables, respectively, and a and c are constants. This model was chosen because it can fit a wide range of asymptotic nonlinear relationships and is not BEAULIEU ET AL.
N2O IN RESERVOIRS
1999
Journal of Geophysical Research: Biogeosciences
10.1002/2015JG002941
Table 2. Thermocline Stratification Index (TSI), Water Temperature, Dissolved Oxygen (DO), Nitrite + Nitrate (NO2.3), Nitrite (NO2 ), Soluble Reactive Phosphorus a + (SRP), and Ammonium (NH4 ) at the Study Sites Reservoir
Date
1
TSI (°C m
Buckhorn Lake (BHR)
8/12/2013
0.9
Barren River Lake (BRR)
8/6/2013
0.5
Brookville Lake (BVR)
8/18/2013
4.1
C.J. Brown (CBR)
8/28/2013
2.1
Caesar Creek (CCK)
9/25/2013
1.9
Carr Creek Lake (CFK)
8/14/2013
1.9
C.M. Harden (CHL)
9/26/2013
2.4
Cagles Mill (CMR) Cave Run Lake (CRR)
10/8/2013 9/5/2013
1.8
East Fork Lake (EFR)
7/24/2013
3.1
8/27/2013
1.4
9/24/2013
0.7
Green River Lake (GRR)
10/29/2013 8/5/2013
NA 2.7
J.E. Roush (HTR) Monroe (MNR)
9/9/2013 8/20/2013
NA 2.5
Mississinewa (MSR)
9/10/2013
1.4
Nolin Lake (NRR)
8/8/2013
2.0
Patoka Lake (PRR)
8/22/2013
4.9
Rough River Lake (RRR)
9/17/2013
0.8
Salamonie (SRR) Taylorsville Lake (TAR)
10/2/2013 9/16/2013
NA 0.8
West Fork (WFR)
8/29/2013
2.2
b
b
b
)
Water Temperature (°C)
Strata epilimnion hypolimnion epilimnion hypolimnion epilimnion hypolimnion epilimnion hypolimnion epilimnion hypolimnion epilimnion hypolimnion epilimnion hypolimnion mixed epilimnion hypolimnion epilimnion hypolimnion epilimnion hypolimnion epilimnion hypolimnion mixed epilimnion hypolimnion mixed epilimnion hypolimnion epilimnion hypolimnion epilimnion hypolimnion epilimnion hypolimnion epilimnion hypolimnion mixed epilimnion hypolimnion epilimnion hypolimnion
28.0 22.4 27.5 22.5 22.9 11.1 23.2 21.1 21.0 12.5 24.8 14.2 22.2 16.3 21.2 26.8 20.0 27.6 17.2 25.6 16.5 22.8 13.5 16.1 26.3 15.1 24.3 25.5 17.8 25.1 22.5 26.8 20.4 25.4 15.5 25.2 20.5 20.3 25.0 18.6 28.0 25.7
+
DO 1 (mg L )
NO2.3 1 (μg N L )
NO2 1 (μg N L )
SRP 1 (μg PL )
NH4 1 (μg N L )
8.1 0.4 9.7 0.3 3.8 0.4 1.0 0.2 4.2 0.3 3.4 0.2 6.2 0.2 2.4 8.6 0.3 5.5 0.2 3.9 0.2 5.7 0.2 0.6 9.0 0.1 6.9 6.3 0.2 8.8 0.5 7.7 0.2 5.4 0.2 5.8 0.3 2.7 4.6 0.3 7.6 0.8
17 19 24 387 1075 1676 6 5 1202 947 13 102 308 20 147 12 11 23 385 5 38 83 32 269 24 263 531 8 7 1784 996 16 526 10 9 13 4 1782 16 13 9 9
2 5 11 55 75 61 5 5 25 6 2 15 50 5 3 6 7 8 69 5 6 33 4 32 10 27 81 2 5 99 23 8 82 3 4 1 1 3 1 2 8 4
23 29 12 27 11 30 16 118 9 29 23 31 9 46 11 18 23 19 94 17 232 30 194 122 13 26 13 15 42 12 44 7 28 8 19 10 28 11 13 274 6 43
8 247 15 142 95 239 258 3458 165 1118 3 64 274 3718 320 6 224 15 99 16 673 173 1256 288 28 256 36 16 892 27 908 11 313 8 372 18 765 15 13 1728 8 622
a
Mean values are presented for the epilimnion and hypolimnion if the reservoir was thermally stratified during sampling; otherwise, the values represent means for the mixed water column. b NA: TSI not calculated for mixed systems.
assumed to represent the underlying mechanisms (supporting information Text S1 and Data Set S1). All data analyses and nonlinear model fitting were conducted in R [R Development Core Team, 2015].
3. Results 3.1. Stratification and Temperature All reservoirs sampled in July, August, and September were thermally stratified, with the exception of HTR (Table 2 and Figure S15a). HTR is among the smallest reservoirs included in the study (Table 1), and 127 temperature profiles collected from the reservoir between May and September of 2000–2010 (J. L. Young, unpublished data, 2013, available from USACE) indicate that the reservoir rarely stratifies, possibly
BEAULIEU ET AL.
N2O IN RESERVOIRS
2000
Journal of Geophysical Research: Biogeosciences
10.1002/2015JG002941
Figure 2. (a–q) Depth profiles of water temperature, dissolved oxygen, and nitrous oxide (N2O) saturation ratio in the deepest location in each of the reservoirs that were sampled during the onetime sampling campaign between July and August of 2013. See Table 1 for a definition of the lake name acronyms listed in each panel.
due to a short water residence time. Overall, the mean epilimnion temperature of the reservoirs sampled between July and September was 25.2°C (range: 21.0–28.°C). The mean hypolimnion temperature was 18.3°C and decreased with increasing reservoir depth (Table 2). The mean TSI value was 1.9°C m1 (range: 0.4–4.9°C m1) for reservoirs sampled during the period of thermal stratification (Table 2). Thermal stratification was completely absent, or the thermocline had sank close to the sediment-water interface, in the three reservoirs sampled in October (CMR, EFR, and SRR), likely because cooler temperatures caused lake waters to circulate after a prolonged period of thermal stratification, a process typically referred to as lake turnover. Depth profiles of nutrient concentrations and water temperature for each lake are presented in the supporting information. 3.2. Dissolved Oxygen Dissolved oxygen was nearly depleted from the hypolimnion of all thermally stratified reservoirs (mean hypolimnion DO = 0.28 mg L1, Table 2). The mean DO saturation in the epilimnion of the stratified reservoirs was 74.6% (Table 2), though DO saturation exceeded 100% at one or more depths in the epilimnion of all but three stratified reservoirs (Figures 2, 4, and S1–23).
BEAULIEU ET AL.
N2O IN RESERVOIRS
2001
Journal of Geophysical Research: Biogeosciences
10.1002/2015JG002941
Table 3. Mixing Status, Presence or Absence of a Middepth N2O Maximum, and Net Effect of Biological Processing on Dissolved N2O Concentration in the Epilimnion and Hypolimnion of Each Reservoir a
N2O Classification Reservoir
Sample Date
Mixing Status
Hypolimnion
Epilimnion
Buckhorn Lake (BHR) Barren River Lake (BRR) Brookville Lake (BVR) C.J. Brown (CBR) Caesar Creek (CCK) Carr Creek Lake (CFK) C.M. Harden (CHL) Cagles Mill (CMR) Cave Run Lake (CRR) Green River Lake (GRR) J.E. Roush (HTR) Monroe (MNR) Mississinewa (MSR) Nolin Lake (NRR) Patoka Lake (PRR) Rough River Lake (RRR) Salamonie (SRR) Taylorsville Lake (TAR) West Fork (WFR)
8/12/2013 8/6/2013 8/18/2013 8/28/2013 9/25/2013 8/14/2013 9/26/2013 10/8/2013 9/5/2013 8/5/2013 9/9/2013 8/20/2013 9/10/2013 8/8/2013 8/22/2013 9/17/2013 10/2/2013 9/16/2013 8/29/2013
stratified stratified stratified stratified stratified stratified stratified fall turnover stratified stratified b mixed stratified stratified stratified stratified stratified fall turnover stratified stratified
East Fork Lake (EFR)
7/24/2013 8/27/2013 9/24/2013 10/29/2013
Reservoir Sampled Repeatedly stratified production production stratified consumption production stratified consumption production fall turnover production throughout water column
Reservoirs Sampled Once consumption production production production production production consumption consumption consumption production production production consumption production production throughout water column consumption production production production production throughout water column consumption production production production production production consumption production consumption neutral production throughout water column consumption neutral consumption neutral
Middepth N2O Maximum Present
x x x x
x
x x
x
a
N2O classification: Classification is based on deviation of dissolved N2O concentration from equilibrium values. Water layers with a mean saturation ratio < 0.80 or > 1.2 were defined as supporting N2O consumption and production, respectively. Intermediate values indicated that biogeochemical processing had a negligible effect on dissolved N2O concentration, and these water layers were classified as neutral. Lakes sampled during fall turnover were not stratified and did not have epilimnion and hypolimnion. These lakes all showed evidence of N2O saturation throughout the water column. b HTR rarely stratifies and was well mixed when sampled.
Dissolved oxygen was undersaturated throughout the water column in the three reservoirs sampled during fall turnover (dissolved oxygen
