Article pubs.acs.org/est
Effect of Water Chemistry and Hydrodynamics on Nitrogen Transformation Activity and Microbial Community Functional Potential in Hyporheic Zone Sediment Columns Yuanyuan Liu,†,⊥ Chongxuan Liu,*,†,‡ William C. Nelson,† Liang Shi,†,§ Fen Xu,†,§ Yunde Liu,†,§ Ailan Yan,†,∥ Lirong Zhong,† Christopher Thompson,† James K. Fredrickson,† and John M. Zachara† †
Pacific Northwest National Laboratory, Richland, Washington 99354, United States School of Environmental Science and Engineering, Southern University of Science and Technology, Shenzhen, Guangzhou 518055, China § School of Environmental Studies, China University of Geosciences, Wuhan, Hubei 430074, China ∥ Institute of Hydraulic and Environmental Engineering, Zhejiang University of Water Resources and Electric Power, Hangzhou, Zhejiang 310018, China ⊥ School of Earth Sciences and Engineering, Nanjing University, Nanjing, Jiangsu 210023, China ‡
S Supporting Information *
ABSTRACT: Hyporheic zones (HZ) are active biogeochemical regions where groundwater and surface water mix. N transformations in HZ sediments were investigated in columns with a focus on understanding how the dynamic changes in groundwater and surface water mixing affect microbial community and its biogeochemical function with respect to N transformations. The results indicated that denitrification, DNRA, and nitrification rates and products changed quickly in response to changes in water and sediment chemistry, fluid residence time, and groundwater− surface water exchange. These changes were accompanied by the zonation of denitrification functional genes along a 30 cm advective flow path after a total of 6 days’ elution of synthetic groundwater with fluid residence time >9.8 h. The shift of microbial functional potential toward denitrification was correlated with rapid NO3− reduction collectively affected by NO3− concentration and fluid residence time, and was resistant to short-term groundwater-surface water exchange on a daily basis. The results implied that variations in microbial functional potential and associated biogeochemical reactions in the HZ may occur at space scales where steep concentration gradients present along the flow path and the variations would respond to dynamic HZ water exchange over different time periods common to natural and managed riverine systems. As, Cr, Hg, Se, Tc, U, and Zn) in groundwater.32−38 Because the HZ is hydrologically dynamic, investigation of the spatial and temporal variations in N transformations under variable fluid flow conditions is essential for understanding biogeochemical processes in the HZ. The transformation of N (Supporting Information (SI) Figure S1) in aquatic environments has been extensively studied with emphasis on denitrification (sequential reduction of NO3− to NO2−, NO, N2O, and N2), dissimilatory nitrate reduction to ammonium (DNRA, NO3− reduction to NO2−, and then to NH4+), nitrification (NH4+ oxidation to NO2−, and then to NO3−), and anammox (anaerobic NH4+ oxidation to N2 in the presence of NO2−).29,39−42 Other processes that can affect N transformation, such as anaerobic nitrification, may
1. INTRODUCTION The hyporheic zone (HZ) is an important ecotone where groundwater and surface water mix, in response to the changes in groundwater or surface water stages due to precipitation, impoundments, injections/withdrawals, tidal influence, or seasonal snow and ice melt.1,2 The HZ is an active biogeochemical region where chemicals and nutrients carried by groundwater and surface water mix and stimulate microbial activities, developing strong chemical gradients that promote rapid transformation of C, N, and other elements.1,3−5 Inorganic N is commonly present in groundwater and surface water and at elevated concentrations is considered a contaminant.6−10 Depending on its oxidation state, N can serve as a microbial electron donor or acceptor11,12 and can mediate other redox reactions.13−16 N transformations in the HZ affect nutrient retention in river systems,17−26 the fate and transport of inorganic N species (NO3−, NO2−, and NH4+), and the emission of N2O.27−31 N redox reactions also influence the subsurface behaviors of other polyvalent elements (Fe, Mn, S, © 2017 American Chemical Society
Received: Revised: Accepted: Published: 4877
October 3, 2016 March 30, 2017 April 8, 2017 April 9, 2017 DOI: 10.1021/acs.est.6b05018 Environ. Sci. Technol. 2017, 51, 4877−4886
Article
Environmental Science & Technology also be important in some environments.14−16,43−47 The transformation of N is affected by water chemistry, the nature of the microbial community, the identity of the N species and their concentrations, dissolved O2 (DO), dissolved and particulate organic carbon (OC), and exchangeable cations and capacity.29,48−59 The transformation of N is also affected indirectly by sediment physical factors such as flow velocity and fluid residence time that affect N, DO, and OC supplies.19,29,54,59,60 In the HZ, the mixing of groundwater and surface water that typically have different chemical compositions can affect microbial community structure and activity, such as net NO3− production and uptake.19,61,62 The net NO3− production and uptake was nonlinearly correlated with fluid residence time in the HZ.19,30 Microbial community structure and functional potential are often characterized by phylogenetic and functional markers (such as rRNA and core housekeeping57,63,64 and key functional genes57,65). Previous studies using these approaches revealed that populations of ammonia oxidizing bacteria57,66 and denitrifiers66−68 in the HZ were correlated with sedimentassociated NH4+ and OC66,67 and water chemistry including O2, pH, NO3−, and dissolved organic N levels.57,67,68 Despite these efforts, the spatial and temporal variations in HZ microbial community functional potential for N transformations was understudied. However, many managed surface water systems, such as those with dams and other hydrologic control, display dynamic stage variations that cause rapidly fluctuating directions of hyporheic exchange, highly varied residence times in the HZ, and transient and temporally variable degrees of groundwatersurface water mixing.69−71 These systems are common, yet poorly studied. Three important research questions exist for these system types: (1) how does microbial community structure and N transformation functional potential respond to dynamic changes in water chemistry and hydrodynamics, (2) how rapid and resistant is this response, and (3) how does this response influence N transformations. We address these three questions using a column system filled with HZ sediments collected from a large, managed waterway (the Columbia River) that experiences significant stage variations (1−3 m) on daily, monthly, and seasonal time frames.72 The column system was subjected to dynamic changes in water flow and chemistry, mimicking observed groundwater and surface water exchange in the field.70,72,73 The primary objective was to determine how HZ microbial community structure and functional potential, and associated biogeochemical processes of N respond to the changes in water chemistry and hydrodynamic conditions. Flow-path sampling of fluids and sediments along the axis of the columns allowed us to determine the impacts of water chemistry and hydrodynamics on the spatial distribution of the microbial community and its influence on N transformation. The results provide new insights into how the spatial and temporal distribution of N transformation and microbial community functional potential may respond to dynamic changes in water chemistry and flow direction in the HZ.
sampling locations and procedures is provided in the SI. The 99% was OC), and 0.12 ± 0.01% total nitrogen (wt/wt, > 99% was organic N), resulting in C:N ratio of 9.9 in molar units. The sediments were stored at 4 °C for 0, 53, and 118 days before they were used to pack Columns A and C, B and D, and E, respectively. During the storage time, inorganic N (including soluble NO3 − and NO 2− and exchangeable NH4+) in the sediment increased (SI Figure S3), indicating that organic N was partially mineralized. The exchangeable NH4+ decreased and soluble NO3− increased, suggesting that nitrification occurred during the storage time at 4879
DOI: 10.1021/acs.est.6b05018 Environ. Sci. Technol. 2017, 51, 4877−4886
Article
Environmental Science & Technology
rate was high (v = 2.20 m·d−1). After the flow rate was decreased at 35 PV (v = 1.21 m·d−1), N2O(g) was higher in the middle of the column (sampling ports B2 and 3). The results indicated that residence time was a key factor controlling relative rates of N2O(g) generation and consumption. 3.3. Effect of Water Chemistry on N Transformation. The concentrations of NO3−, NO2−, N2O, and NH4+ in the effluents from Columns A and B were higher than those from Columns C and D because SGW contained a higher NO3− concentration than SRW (Figures 1a−d) as observed in the field. After eluted out of Columns A and B, the effluent NO3− concentration slowly increased with time, indicating a decrease in NO3− reduction rate with time. This rate decrease was attributed to the consumption of labile OC through microbial activity and transport and export of OC from the column system. OC is the primary electron donor and carbon source for denitrification and DNRA, and a potential source of NH4+. OC was detected in the effluent, indicating that some was transported out of the column during the experiment (SI Figure S7). The variation in effluent NO3− concentration for Column D (Figure 1d) over the first 35 PV, however, cannot be fully explained by labile OC loss through microbial degradation and/ or DOC export. The effluent NO3− concentration increased rapidly over the first 20 PV, and then reached a plateau of approximately 0.036 mM. After the flow rate was decreased at 35 PV, NO3− decreased dramatically, and then increased at a much slower rate after approximately 70 PV. Moreover, effluent NH4+ decreased when effluent NO3− reached the 0.036 mM plateau. These results suggested that the NO3− concentration was also affected by the reactive transport of DO. Although DO was not detected in the effluent of Column D; it was detected in upgradient pore water at the sampling ports (SI Figures S6b4). As DO migrated into the column, NO3− reduction was suppressed and NH4+ was oxidized to NO3− (SI Figure S6b1), resulting in a rapid increase in effluent NO3−. The decrease in flow rate and increase in residence time apparently enabled DO consumption before it reached the second or downgradient half of the column. We interpret this as denitrification and DNRA dominating in the second half of the column, resulting in lower NO3− and higher NH4+ concentrations in the effluent. The changes in NO2− and N2O concentrations in the pore water samples were consistent with the changes in denitrification and DNRA rates: they were lower at the inlet or upgradient end of the column where denitrification and DNRA were suppressed (SI Figure S6b2 and b3). 3.4. N Transformation under Dynamic GroundwaterRiver Water Exchange. In Column E, SGW and SRW were switched every 2−3 PV from opposite ends of the column: the effluent side became the inlet side during the switch, and vice versa (Figure S2). At the beginning of each new injection period except for the first one, the effluent solution was from the previous injection period beginning first with drainage from the upgradient end of the column. As a result, a rapid decrease in effluent N species and DO was observed within the first PV (Figure 1e). The effluent NO3− at the beginning of the SGW injection period was higher than the influent NO3− in SRW because of nitrification during the previous SRW injection period. This observation was consistent with measurements on column sediments after the last SRW injection period, where soluble NO3− near the SRW inlet (equivalent to 0.155 mM in pore water) was higher than the influent NO3− concentration (0.044 mM). After drainage from the previous injection period,
4 °C (SI Figure S3). Functional genes for denitrification, DNRA, nitrification, and anammox were all identified in the sediments (SI Figure S4b), implying that the microbial community in the HZ has extensive capabilities for inorganic N transformations. 3.2. Effect of Residence Time on N Transformation. Br− transport displayed slightly early breakthrough in all columns (SI Figure S5), where the relative concentrations of C/C0 = 0.5 were attained before 1 pore volume (PV). The results implied the presence of preferential flow paths in the columns.83,84 NO3−, NO2−, N2O, and NH4+ were detected in the effluents of Columns A through D (Figures 1a−d, left-hand data panels). When the flow rate was low and the residence time high (Columns A and C, v = ∼ 0.5 m·d−1), the effluent profiles of NO3−, NO2−, and N2O showed a delayed breakthrough with lower concentrations as compared to the faster flow columns (Columns B and D, v > 1.1 m·d −1 ). Effluent NH 4 + concentration was higher at a lower flow rate because the longer residence time promoted DNRA activity and/or organic N mineralization with NH4+ as the end products. Effluent NO2− was a likely product of either NO3− reduction or NH4+ oxidation by DO. Mn-oxides were ruled out as a possible oxidant because of their low abundance in the Hanford HZ sediments.70,85 Soluble NO3− and NO2− and exchangeable sediment NH4+ were quantified after the experiments were completed (Figures 1a−d, right-hand data panels). Soluble NO3− rapidly decreased at the inlet side of slower flow columns (Columns A and C). The soluble NO3− near the inlet of Column A was 0.02 μmol· g−1 (equivalent to 0.055 mM in pore water), which was much lower than the influent NO3− concentration (0.447 mM), indicating rapid NO3− reduction through denitrification and/or DNRA. The reduction of NO3− by assimilatory nitrate reduction is unlikely in the presence of NH4+ since the latter would be energetically favored.86 Similarly, the soluble NO3− concentration in Column C rapidly decreased from 0.028 μmol·g−1 (at −0.2 cm) to 0.010 μmol·g−1 (at −7.2 cm) at the inlet side of the column. In contrast, soluble NO3− gradually decreased from the inlet to outlet in Column B; whereas it did not decrease appreciably in Column D. Soluble NO2− was low in all columns except near the inlet of Column A, where rapid NO3− reduction likely led to the accumulation of NO2−. Exchangeable NH4+ was lower in sediments near the column inlets, then it either increased gradually toward the outlet of the column (Columns B−D) or stabilized after reaching a plateau of approximately 2 μmol·g−1 (Column A). The net differences in exchangeable NH4+ from the initial to the end of the column experiments were >1.3 μmol·g−1 in the slower flow columns (Columns A and C) and 0) and production (
