http://orcid.org/0000-0002-8426-5502
Figures
Abstract
Nitrate pollution of surface and groundwater systems is a major problem globally. For some time now wetlands have been considered potential systems for improving water quality. Nitrate dissolved in water moving through wetlands can be removed through different processes, such as the denitrification process, where heterotrophic facultative anaerobic bacteria use
for respiration, leading to the production of nitrogen (N2) and nitrous oxide (N2O) gases. Nitrate removal and emission of N2O in wetlands can vary spatially, depending on factors such as vegetation, hydrology and soil structure. This study intended to provide a better understanding of the spatial variability and processes involved in
removal and emission of N2O in riparian wetland soils. We designed a laboratory experiment simulating surface water flow through soil columns collected from different sites dominated by different plant species within a wetland. Water and gas samples for
and N2O analyses were collected every 5 days for a period of 30 days. The results revealed significant removal of
in all the soil columns, supporting the role of riparian wetland soils in removing nitrogen from surface runoff. Nitrate removal at 0 and 10cm depths in sites dominated by Phragmites australis and Carex schnimdtii was significantly higher than in the site dominated by Calamagrostis epigeio. Nitrous oxide emissions varied spatially and temporally with negative flux observed in sites dominated by P. australis and C. schnimdtii. These results reveal that in addition to the ability of wetlands to remove
, some sites within wetlands are also capable of consuming N2O, hence mitigating not only agricultural nitrate pollution but also climate change.
Citation: Mwagona PC, Yao Y, Yuanqi S, Yu H (2019) Laboratory study on nitrate removal and nitrous oxide emission in intact soil columns collected from nitrogenous loaded riparian wetland, Northeast China. PLoS ONE 14(3): e0214456. https://doi.org/10.1371/journal.pone.0214456
Editor: Debjani Sihi, Oak Ridge National Laboratory, UNITED STATES
Received: June 30, 2018; Accepted: March 13, 2019; Published: March 28, 2019
Copyright: © 2019 Mwagona et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files.
Funding: This study was supported by The Fundamental Research Funds for the Central Universities (2572017CA15) and Natural Science Foundation of Heilongjiang Province (QC201604). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Nitrate pollution of surface and groundwater systems has become a major problem globally. Both human activities and natural cycles are sources of
; however, human activities, such as the use of organic and inorganic fertilizers, poor placement of livestock waste, and effluent from untreated sewage, have been documented as major sources of
pollution [1]. Use of nitrogenous fertilizer to enhance agriculture production has increased tremendously in the recent past to meet the food demand of the rapidly growing population. Unfortunately, high concentrations of
in drinking water can cause serious health problems in humans, such as methemoglobinemia in infants and young children [2, 3]. In addition, recent studies have indicated increasing risk of stomach cancer and neural tube defects as a result of drinking water with a high concentration of
[4]. Moreover, excess
can cause environmental problems by inducing eutrophication in aquatic systems [5]. Despite having a long history of water system management, China is currently facing a serious problem of
pollution in its aquatic systems hence risking the health of billions of people. While assessing
pollution of groundwater in agricultural dominated area northern china, [6] reported
concentrations of 50 mg/L
exceeding the allowable limit in drinking water.
For some time now wetlands have been considered potential systems for improving both surface and ground water quality [7, 8]. Nitrate in polluted water moving through wetlands can be removed through different processes, such as denitrification. These processes can vary spatially, depending on factors, such as vegetation, hydrology, organic matter and soil structure [9, 10]. Biological denitrification is one of the most important processes removing inorganic nitrate in wetlands under anaerobic conditions where several bacteria are involved [11, 12]. Under low dissolved oxygen concentrations, denitrifying bacteria use in respiration in the presence of carbon sources such as organic matter. However, this can lead to emission of nitrous oxide (N2O), a notable greenhouse gas species as an intermediate product and dinitrogen (N2) as the end product. The global warming potential of N2O is about 340 times compared with carbon dioxide (CO2), and it is also destructive to stratospheric ozone [13]. Availability of
carbon and other environmental factors such as temperature, pH, soil moisture and biological denitrification rates greatly influence emission of N2O [14]. Studies show that wetlands loaded with
produce more N2O compared to non
loaded wetlands. For instance [15] noted that a nitrate-loaded riparian forested buffer zone produced up to 20 kg N /ha /yr of N2O compared to non
loaded grassland buffer zone 2–4 kg N /ha /yr of N2O.
Qixing River Wetland National Nature Reserve is one of the most important and a typical representative remnant of an inland freshwater marsh type located in Sanjiang plain, the agricultural hub of China. The wetland is located in the middle of farmlands and it receives agricultural runoff rich in nitrogenous fertilizers from all directions. Qixing River Wetland National Nature Reserve is characterized by high environmental heterogeneity with some sites purely dominated by mono-stands of C. epigeios, P. australis and C. schnimdtii plants. Although studies have shown that NO3- removal capacity and greenhouse gas production from wetland soils can vary spatially and temporally [9, 16], it is not clear whether this occurs at the Qixing River Wetland National Nature Reserve, despite its heterogeneous nature and high nitrogeneous fertilizer loading. A better understanding of the spatial variability and processes involved in removal and N2O emission is very crucial for the management and restoration of degraded sites and reestablishing wetlands to reduce pollution within Sanjiang Plain. In addition, the findings of this study expound on the limited scientific papers published on nitrogen removal efficiency from a wetland loaded with high nitrate [9].
To our knowledge no studies have been conducted in Qixing River Wetland National Nature Reserve to assess the spatial variability in attenuation of and emission of N2O. The objective of this study is to determine spatial variation of
attenuation and N2O emission using intact soil columns collected from sites dominated by different vegetation types. We hypothesized that
removal capacity and N2O production from the riparian wetland would vary spatially and with depth because of the high environmental heterogeneity creating different micro-environments within the vegetation types. To unravel this, soil columns collected from sites dominated by different vegetation types (C. epigeios, P. australis and C. schnimdtii) were incubated in the laboratory, treated with
enriched water and measured for the concentrations and fluxes of different N species at varying soil depth, which could eventually infiltrate and become groundwater in a natural setting.
Study area and sampling procedure
Study area
The study was conducted in Qixing River Wetland National Nature Reserve, with permission from the National Nature Reserve administration, which is responsible for the protection of the reserve. Qixing River Wetland National Nature Reserve is a typical and representative riparian natural freshwater marsh type located in Sanjiang Plain, Northeastern China (46°39’45”-46°48’24”N, 132°00’22”-132°24’46”E) at an average elevation of 80m (Fig 1). The reserve is approximately 20,000ha in size and it is divided into a buffer zone (3,600ha), core zone (7,960ha) and experimental zone (8,440ha). This study area is under the influence of temperate humid monsoon climate with an average yearly temperature and rainfall of about 1.9°C and 550 mm, respectively. Qixing River Wetland National Nature Reserve soil type is characterized by mires soils, including peat-boggy soils, humus, meadow and peat soils.
Bottom at the right corner is the location of Northeast China (A) and shaded grey and yellow dot is Sanjiang Plain and Qixing River Wetland National Nature Reserve, respectively. The enlarged map at the top (B) is Qixing River Wetland National Nature Reserve and 1, 2, 3 are the sampling sites.
The reserve is one of the most preserved, almost virgin and important ecosystems in the Sanjiang Plain, supporting a diversity of flora and fauna. It provides a habitat for about 29 threatened species of which 26 are birds and 3 are mammals. Moreover, the wetland is endowed with diverse species of plants and the most dominant ones are C. epigeios, P. australis and C. schnimdtii. Other companion species include Quercus mongolica, Glyceria spiculosa, Betulla platyphylla among many others. Apart from providing ecological functions, Qixing River Wetland National Nature Reserve also maintains good water quality and provides water storage services, thereby allowing infiltration to occur, hence recharging aquifers and groundwater. In addition the wetland provides significant socio-economic and cultural benefits.
In the recent years, wetlands that were once widely distributed in the Qixing catchment has become patches embedded in agricultural landscape due to long-term agricultural development [17]. The remaining Qixing River Wetland National Nature Reserve is surrounded by agricultural farmlands (Fig 1). The farmlands have been converted from drylands to paddy field. Due to agricultural activities, the wetland is strained in offering its environmental functions and services.
Soil sampling
Three sites were selected based on vegetation types; C. epigeios, P. australis and C. schnimdtii within the wetland experimental zone (area reserved for research) to allow for comparison. During the sampling period, the standing water depth in sites dominated by P. australis and C. schnimdtii ranged approximately from 0.4 to 0.7m and from 0 to 0.2m in the site dominated by C. epigeios. At every sampling site, three intact soil cores, excluding the plants, were collected using polyvinylchloride (PVC) pipes (4.5 cm internal diameter, 100cm in length) drilled with holes on its side to be used for collecting water samples in the laboratory. The holes were equally spaced at a 10cm interval (0, 10, 20 and 30 cm). The PVC pipes were sharpened at the bottom for easy penetration into the sediment during drilling, as described by [9]. The pipes were manually drilled to a depth of 40cm into the soil. Immediately after retrieving every PVC pipe with the soil content inside, both ends were sealed with cork and the sampling holes were covered with waterproof tape to avoid water leakage. Soil samples for determination of total nitrogen (TN), total organic carbon (TOC) and bulk density were collected at every sampling site following the same procedure. The samples were then transported to the laboratory and stored at room temperature for three to five days before the experiment was set up and sampling occurred. Soil samples for determination of TN and TOC were sectioned to get nutrients for each level.
Laboratory experiment setup
In the laboratory, the columns were set up to simulate downward surface water flow through the soil columns collected at different sites. The experiment consisted of two reservoirs A and B positioned at different heights on the same stand (Fig 2). Reservoir A was filled with nitrate solution (60 mg/L ) prepared in the laboratory by dissolving potassium nitrate in distilled water. The solution from reservoir A flowed to reservoir B due to the air pressure generated when the solution level in B reduced below the nitrate solution level of the tube. The tube in reservoir B was used to regulate the nitrate solution level. The waterproof tape seal on the holes was replaced with Rhizon soil moisture samplers that were used to collect water samples from the soil using a nylon syringe. All the nine PVC pipes (three from each site) were connected to a common supplier of nitrate solution from the smaller reservoir B, and the flow rate was set at 0.042mL/second using a valve. The inlet from reservoir B into the columns was placed 5cm above the 0cm depth (soil-water surface) in all the cores (Fig 2). To ensure a constant water head in each soil column during the experiment, the water level in reservoir B was the same as in the inlet at any given time to allow the nitrate solution to percolate through each soil column.
Gas sampling
Prior to running the experiment, initial gas samples were collected from the column headspace and injected in aluminum-multi-layer foil composite film gas sampling bags. This was done by closing the PVC pipe ends with rubber stoppers and the gas was collected from the head space volume at 0, 10, 20 and 30 minutes interval using 20 ml nylon syringes. The gas samples were then stored into the dark at room temperature for 3 days before analysis. After 30 minutes, the rubber stoppers were removed and the experiment was left to run. Subsequent sampling was done every 5 days following the same procedure. The N2O gas concentration was measured using a gas chromatograph (Agilent 7890A, Agilent Co., Santa Clara, CA, USA) equipped with an electron capture detector (ECD). The slope of the N2O gas concentration over time (μmol/mol/h) during sampling was used to compute the N2O flux(g/m2/h) between soils and the atmosphere according to the equation from [18] and [19]
(1)
where dc/dt is the slope of the N2O gas concentration curve variation, along with time. M (g/mol) is the molar mass of N2O gas. P (kPa) is the atmospheric pressure in the sampling site. T (°C) is the absolute temperature during the sampling. V0, T0, P0 are the N2O gas molar volume, air absolute temperate and atmospheric pressure under standard conditions, respectively. H (m) is the height of the PVC column during sampling. Positive flux values mean the soil is a source of N2O to the atmosphere, and negative values mean the soil is a sink of N2O.
Water sampling
In order to allow the system to attain equilibrium with the water flow rate, the experiment was run for 5 days before the initial sampling took place. After that, sampling was done at 5 days intervals for a period of 30 days from 4 different depths (0 (soil-water surface), 10, 20 and 30cm (lower outlet)) using a 20 ml nylon syringe attached to the Rhizon soil moisture sampler. Water samples were also collected at the inlet (5cm above the soil-water surface) for every sampling occasion. Only 12 ml of water was sampled for every depth on every sampling occasion. The water samples for and ammonium (
) were immediately filtered through 0.45μm GD/X Whatman filters and stored at -18°C until analysis. Nitrate and
were analyzed using standard methods as described in [20] and [21] respectively. Nitrate removal in percentage (Nr) was calculated using the following equation
(2)
where IC is the concentration of
(mg N/L) at the inlet and Cd is the concentration of
(mg N/L) at a given depth.
To determine nutrients fluxes, the formula proposed by [9] was used. First, the mean values of nutrient concentrations for the replicate columns were determined from each depth as follows:
(3)
where
= the average of the nutrient concentration X (mg N/L) determined from the site j in the soil column c at depth d on day i, while n is the number of soil columns collected from each site. Then the average flux (F) of nutrients in the soil columns and at different depths was calculated using the measured nutrient concentrations and flow rates.
Where is the average nutrient concentration (mg/L) in the outflow and
is the average nutrient concentration (mg/L) in the inlet (5cm above soil-water surface) [22].
Statistical analyses
Before analysis, all the variables were tested for normality using the Kolmogorov–Smirnov test and homogeneity of the data was tested using the Bartlett test. With exception of pH all the variables were log transformed to satisfy the normality and variance assumption before conducting analysis. For the different soil profiles (0, 10, 20 and 30cm depth) repeated measure ANOVA was carried out to determine the differences in concentration among the sampling sites and depth. For multiple comparisons in case of a significant ANOVA result, post hoc comparisons using the Tukey HSD test were applied. Multiple regression analyses were performed using the forward stepwise selection procedure to select those predictor variables with a significant F-value that significantly increased the regression sum of squares. The predictor and response variables used are pH, TOC, TN, bulk density and C/N and
and
, respectively. Figures were drawn using R software (version 3.4.1) [23]. Note that the results were reported as mean value ± standard error (SE).
Results
Soil characteristics
The soil collected from the three sites within the Qixing River Wetland National Nature Reserve had distinct characteristics as indicated in Table 1. Overall, total nitrogen (TN) and total organic carbon (TOC) concentration decreased with increasing depth (from 0-40cm) while bulk density increased with increasing depth. Average soil total organic carbon computed between 0-40cm depths was 27.14, 18.29 and 17.60 mg C/g at the site dominated by C. epigeios, P. australis and C. schnimdtii, respectively. Similarly, mean total nitrogen was 2.41, 2.26 and 2.65 mg N/g at sites respectively. Mean soil bulk density for the sites dominated by C. epigeios, P. australis and C. schnimdtii was 1.12, 1.04 and 1.03 g cm-3 respectively and could be categorized as fine silt and clay soil. Notable are the low values of bulk density less than 0.5 g cm-3 from 0-10cm depth at the sites dominated by P. australis and C. schnimdtii indicating that the soil is very rich in organic matter. Soil pH was relatively low in the site dominated by C. epigeios.
Results are in mean ± SE.
Reduction of nitrate in different soil depth
The mean concentration of for the 30 days study period at the inlet (5cm above soil-water surface) in the soil columns ranged between 59.32 and 59.37 mg N/L while the mean lower outlet (30cm) concentration ranged between 0.59 and 0.97 mg N/L (Table 2). This indicates that
concentrations were reduced with increasing depth in all the soil columns. At the 0cm (soil-water surface) depth, sites dominated by P. australis and C. schnimdtii had a higher
reduction of about 60% which is five times higher compared to that of C. epigeios at the same depth. At 10cm depth mean
concentration had reduced to 20.06, 4.26 and 0.84 mg N/L for sites dominated by C. epigeios, P. australis and C. schnimdtii, respectively which corresponded to 66.20, 92.82 and 98.58%
removal respectively (Fig 3). With the exception of the site dominated by C. epigeios, the
concentration for the other two sites was almost constant at 20 and 30cm (lower outlet) depths (Fig 3). Ammonium concentrations measured at the inlet ranged between 0.59 and 0.60
mg N/L and increased between inlet and outlet for all soil columns (Fig 4). The rate of
mg N/L increase was remarkably higher at 0cm depth in soil columns collected from the site dominated by P. australis followed by C. schnimdtii.
Results are in mean ± SE for the whole study period (30 day time period).
In view of the fluxes of (g N/m2/h), negative values (nitrate consumption) were recorded in all different depths of the three soils columns revealing that the soils are net sink (Fig 5). There was statistical significantly difference in
influxes between soil depth in all sites as determined by one-way ANOVA (p<0.05). A post hoc test revealed that
influxes at the 0cm depth in sites dominated by P. australis and C. schnimdtii differed significantly from the other depths (i.e. 10, 20 and 30cm). On the other hand,
fluxes at 20 and 30cm depths did not show statistical significant difference at the site dominated by C. epigeios. The pattern of
fluxes (mg N/m2/h) differed from
fluxes in all depths of all the soil columns. Positive fluxes of
were recorded in all depths of all three soils indicating that the soils are net sources (Fig 6). Mean fluxes of
in sites dominated by P. australis and C. schnimdtii differed significantly within soil depths p = 0.014 and p = 0.034, respectively. Post hoc test revealed that at the site dominated by P. australis, the mean
fluxes were significantly lower at 20cm depth (21.67±3.30 mg N/m2/h, p = 0.027) and at 30 cm depth (63.09± 17.46 mg N/m2/h, p = 0.041) compared to 0cm depth (500.49±170.64 mg N/m2/h). Similarly, post hoc test indicated that the mean efflux of
at 0 cm depth in site dominated by C. schnimdtii was statistically significantly higher than at 30cm depth (p = 0.047).
fluxes were not statistically significantly from the different depths in site dominated by C. epigeios.
(a) Soil from site dominated by C. epigeios, (b) soil from site dominated by P. australis, (c) soil from site dominated by C. schnimdtii. Error bars represent standard error of the mean.
(a) Soil from site dominated by C. epigeios, (b) soil from site dominated by P. australis, (c) soil from site dominated by C. schnimdtii. Error bars represent standard error of the mean.
Temporal variation of 
concentration in different depth
Figs 7–9 shows the temporal variation of concentration from the different depths of the soil column experiment over the 30 day period. Generally, the concentration of
decreased remarkably in the first 5 days of the experiment. For the soil column from the site dominated by C. epigeios,
concentration at the 0cm depth (soil-water surface) was 43.22 mg/L corresponding to 27.19%
removal in the first 5 days from the initial concentration at the inlet of 59.36 mg/L (Fig 7). Then it increased from 43.22 mg/L to an average of 51.67 mg/L for the remaining days of the experiment corresponding to 12.95% of
removed. For sites dominated by P. australis and C. schnimdtii 60%
was removed in the first 5 days at the 0cm depth (soil-water surface) (Figs 8 and 9). Further rapid decrease in
concentration was observed from the site dominated by C. schnimdtii on day 10 followed by a gradual increase in
concentration up to day 30. Unlike at the 0cm depth (soil-water surface) which showed a significantly temporal variation of
concentration among the soil columns, the
concentration patterns in the other depths (10, 20 and 30cm (outlet)) were almost similar for the C. schnimdtii (Fig 9). After the first 5 days of the experiment, the concentration of
for C. schnimdtii increased gradually at 0cm (soil-water surface) depth until it reached a peak on the 20 day, followed by a slight decrease. This clearly shows that
removal was very high during the first 5 days in the soils collected from the site dominated by C. schnimdtii.
Nitrous oxide emission from soil column
The N2O fluxes measured from the soil columns are presented in Fig 10. Overall, the N2O emission depicted temporal and spatial variation during the sampling period. Hourly, mean N2O emission from the site dominated by C. epigeios, P. australis and C. schnimdtii ranged between 0.76 to 13.93 μg m-2 h-1, -1.25 to 12.14 μg m-2 h-1 and -3.40 to -0.35 μg m-2 h-1 respectively (Fig 10). Assuming a constant N2O fluxe over the year, the annual fluxes would be between -297.84 to 1220.27 g N2O ha-1 y-1 for the three sites dominated by different vegetations within the wetland. The soil from the site dominated by C. epigeios was a net source of N2O, while soils from sites dominated by P. australis and C. schnimdtii were net sinks of N2O. Nitrous oxide emission from the site dominated by C. epigeios increased gradual from day 1 until it reached on a peak on day 20 and decrease afterward (Fig 10A). Similar but reverse pattern to that from the site dominated by C. epigeios was observed in soil from the site dominated by C. schnimdtii reaching a peak of N2O consumption on day 20 (Fig 10C).
(a) Soil from site dominated by C. epigeios, (b) soil from site dominated by P. australis, (c) soil from site dominated by C. schnimdtii. Error bars represent standard error of the mean.
Table 3 shows the correlation results between pH, TOC, TN, C/N, concentration and
concentration. From the results
depicted a significantly positive relationship with TOC and TN. A multiple forward stepwise regression revealed that TOC was the most important variable related to
removal (
removal = -9.310+0.984TOC, p = 0.000, R2 = 0.77). On the other hand,
removal was significantly related with pH and TN with a multiple forward stepwise regression revealing pH as the most important variable (
= -5.592+1.27pH, p = 0.009, R2 = 0.52)
Discussion
Spatial dynamics of nitrate and ammonium in wetland soil
Wetlands have the ability to improve surface water quality by reducing pollution. Previous studies have revealed plant uptake, soil retention, denitrification, microbial immobilization and nitrification as the major process of nitrogen retention in wetlands [24–26]. In wetlands which receive high nitrate loading from agricultural runoff or sewage treatment plant discharge, denitrification has been found as the most dominant process of reducing
[26, 27]. The results of our experiment revealed remarkable reductions in
concentrations from between 59.32 and 59.37 mg N/L at the inlet (5cm above soil-water surface) to concentrations below 1 mg N/L at the lower outlet of the soil columns. Therefore, almost 98% of the
disappeared as the water flowed down through the soil columns indicating an efficient nitrate removal capacity of the wetland soils. These findings are in agreement with other studies which have confirmed nitrate removal efficiency as high as 90% in nitrate-loaded riparian wetlands [9, 28]. Similar observations were also made in constructed wetlands systems. For instance, [29] reported significantly nitrate removal efficiencies of about 70–99% in a small scale flow constructed wetland system. Denitrification which has been documented as a
sink in riparian areas by researchers [9, 30], was probably the major pathway of nitrate removal; however, dissimilatory nitrate reduction to ammonium (DNRA) also might have played a role. Actually, the increasing concentration of
between the inlet and outlet implied the existence of DNRA or mineralization of organic matter. Studies have shown that in the presence of high soil organic carbon, DNRA can occur jointly with denitrification under reducing environments [9, 31–33]. This concur with the findings of this study as the production of
was high at 0cm (soil-water surface) and 10cm depths where the total organic carbon was also high. Note, however, that
concentrations do not support DNRA as the dominant process responsible for reducing
in the soil columns as the amount of N accumulation as
does not account for the N lost as
.
From our results, it is quite clear that the soils collected from different sites within the wetland acted as a sink for . On average, the
fluxes were 4.95, 4.91 and 3.78 g N/m2 in the sites dominated by C. schnimdtii, P. australis and C. epigeio, respectively. These values are almost in the same range as those reported from laboratory studies in Danish peat soil (4.6 g N /m2) by [34], but much higher than those reported in laboratory microcosms in central and northern Jutland, Denmark (1.1 g N/m2) by [9].
Although the average removal rate from the inlet to the lower outlet was almost the same from the three sites, a significant vertical gradients of
removal rates along the soil profiles were evident with the highest rates occurring in the upper layers. The differences in
removal rates along the soil profiles were most likely the result of differences in total organic carbon. The fact that there was no significant difference between
fluxes at 20 and 30cm depth in the soil columns suggests that most of
removal in this system occurred at depths above 20 cm where total organic carbon was also very high. As indicated in Table 1, all the soils used in this study contained less than 5% TOC while the
removal rate was very high. This agrees with previous studies showing that
removal was not C limited in a nitrogen loaded riparian wetlands with a TOC content less than 7% [9]. Upper soil profiles contained the highest amounts of organic carbon (Table 1). This observation was further supported by the result of multiple linear regression which singled out TOC as important variable related to
removal (Table 3). [27] and [30] revealed that
removal through the process of denitrification in riparian wetland soil is very high in the upper layers and depends on microbial activity and soil organic carbon level. [35] further observed a significant relationship between
reduction and extractable carbon. Interestingly, the results of our study further revealed variation in
removal at 0cm (soil-water surface) and 10cm depths. In the sites dominated by P. australis and C. schnimdtii,
removal was significantly higher than in the site dominated by C. epigeio (Fig 3). This is contrary to the expectation, since the upper soil profiles of the site dominated by C. epigeio had higher amounts of total organic carbon compared to the other sites (Table 1).
This spatial variability in may be explained by organic carbon quality. Scores of studies reported that removal of
through denitrification process is highly correlated with available carbon (e.g. microbially labile) rather than total organic carbon [27, 36, 37]. Thus, although C. epigeio had higher concentrations of total organic carbon, the available carbon was potentially low hence reducing
removal. Note, however, that available carbon (e.g. microbially labile) was not measured in this study. While assessing the dynamics of microbial communities during decomposition of plant litter from soil ecosystems, [38] observed that high C/N ratios of C. epigeio litter is less attractive for microbial degraders. Therefore it is also possible that the high C/N ratio observed in the upper layers (0-10cm) of the site dominated by C. epigeio may have affected the removal rate of
.
The variability in removal could also be due to soil pH and denitrifying enzymes sensitivity. Soil pH is one of the essential factors that can control denitrification process through enzyme sensitivity [39, 40]. Low soil pH, as observed in the site dominated by C. epigeio could limit the availability of organic carbon and available mineral N to denitrifying bacteria [39, 41]. However, this should be interpreted with caution because scientists have argued differently on the relationship between pH and the denitrification process. For instance [42] observed that below pH 6 denitrification was somewhat retarded but even at pH 4.9 more than 70% of added nitrate was lost within 2 weeks. On the other hand, [43] pointed out that the rate of carbon mineralization (measured as CO2 production), rather than pH, controlled the rate of denitrification in a system.
Nitrous oxide emission from wetland soil
Previous studies have reported temporal and spatial variability in N2O fluxes from natural riparian wetlands [44, 45]. Some riparian wetland soils have been found to act as sources of N2O [46] and others as sinks of N2O [45]. In the present study, the emission of N2O varied spatially and temporally. The average N2O fluxes measured in the sites dominated by C. epigeios, P. australis and C. schnimdtii of 0.76 to 13.93 μg m-2 h-1, -1.25 to 12.14 μg m-2 h-1 and -3.40 to -0.35 μg m-2 h-1, respectively are almost in the same range to those reported by [45] from four Danish riparian wetlands of -44 to 122 μg N2O -N m-2 h-1. We hypothesized that N2O production from the riparian wetland would vary spatially because of the high environmental heterogeneity that created different micro-environments within the vegetation types. Our findings clearly supported our hypothesis. Positive N2O fluxes were observed from soil collected from the site dominated by C. epigeio, implying production of N2O by the soil to the atmosphere. On the other hand, soils collected from sites dominated by P. australis and C. schnimdtii had negative fluxes of N2O implying consumption of the gas by the riparian wetland soil. These differences could potentially be ascribed to: (i) differences in total organic matter. In their study, [47] revealed that sites loaded with high total organic carbon in boreal upland soils were net sources of N2O while low total organic carbon sites were net sinks of N2O. This is in agreement with our findings, since the average soil organic carbon from the site dominated by C. epigeio was higher compared to that of sites dominated by P. australis and C. schnimdtii, and (ii) removal efficiency, from the results (as shown in Figs 3 and 5), the relative higher
removal efficiency in the upper layers of P. australis and C. schnimdtii compared to C. epigeio probably could have stimulated N2O uptake during denitrification [45, 48, 49]. On the other hand, the high
concentration on the upper layers of C. epigeio could have hindered N2O reduction during denitrification [45, 50]. Note that there is no substantial increase in
under C. epigeio to attribute N2O production to DNRA.
Soil pH is also one of the most important factors that affect denitrifier community composition, denitrification rates and denitrification end products [39]. Our results revealed relatively high positive fluxes in the site dominated by C. epigeio where the soil pH values were relatively low (Table 1 and Fig 10A). These findings are in agreement with [51] and [52] who reported that low soil pH inhibits di-nitrogenoxide reductase thus increasing the amount of N2O production. On the contrary, [39] and [45] found that low soil pH (<5) was associated with low N2O production. It is quite clear that the influence of low soil pH on N2O production is still debatable and possibly inconsistent. While assessing the relationship between N2O production and soil pH, [53] noted that the net effect of low pH on N2O is not straightforward since multiple environmental factors can have influence on denitrification rates. Note, also, that the generally low N2O emission observed in this study could also be explained by the use of laboratory soil columns which allows small-size areas to be sampled and monitored hence possibly missing hotspots. Moreover, the sampling time and frequency may have missed peak hot moments of N2O production [54].
Conclusion
This laboratory experimental study was intended to provide a better understanding of the spatial variability and processes involved in the attenuation of and the emission of the greenhouse gas N2O in riparian wetland soils located in the middle of farmlands. The riparian wetland receives agricultural runoff rich in nitrogenous fertilizers from all directions. In the laboratory, the experiment was designed to simulate downward surface water flow through the soil columns collected at different sites dominated by different vegetation types (C. epigeios, P. australis and C. schnimdtii). Our results revealed significant and rapid removal of
in all of the soil columns, supporting the crucial role of riparian wetland soils in removing N from surface runoff. Our study further revealed that
removal at 0cm (soil-water surface) and 10cm depths in the sites dominated by P. australis and C. schnimdtii was significantly higher than in the site dominated by C. epigeio. This could be attributed to organic carbon quality and low pH values in the site dominated by C. epigeio. Moreover, our study showed that N2O emissions varied spatially and temporally with negative flux observed in soils from the sites dominated by P. australis and C. schnimdtii. It is clear from this study that in addition to riparian wetland soil being able to remove
, some sites within the wetland are capable of consuming N2O hence mitigating not only health problems (e.g. methemoglobinemia in infants and young children) and environmental problems (e.g. eutrophication) related to
but also atmospheric N2O pollution.
Acknowledgments
We are grateful to all the people who helped this project along the way. Without them, this paper, and the research that accompanied it, would have been considerably more difficult, if not impossible. The authors sincerely acknowledge the anonymous reviewers for their thorough revision and invaluable comments which have helped to significantly improve this paper.
References
- 1. Della Rocca C, Belgiorno V, Meriç S. Cotton-supported heterotrophic denitrification of nitrate-rich drinking water with a sand filtration post-treatment. Water Sa. 2005;31(2):229–36.
- 2. Fan AM, Steinberg VE. Health implications of nitrate and nitrite in drinking water: an update on methemoglobinemia occurrence and reproductive and developmental toxicity. Regulatory toxicology and pharmacology. 1996;23(1):35–43. pmid:8628918
- 3. Magee PN, Barnes J. The production of malignant primary hepatic tumours in the rat by feeding dimethylnitrosamine. British journal of cancer. 1956;10(1):114. pmid:13342328
- 4. Ward MH. Workgroup report: drinking-water nitrate and health—recent findings and research needs. Environmental health perspectives. 2005;113(11):1607. pmid:16263519
- 5. Smolders AJ, Lucassen EC, Bobbink R, Roelofs JG, Lamers LP. How nitrate leaching from agricultural lands provokes phosphate eutrophication in groundwater fed wetlands: the sulphur bridge. Biogeochemistry. 2010;98(1–3):1–7.
- 6. Zhang W, Tian Z, Zhang N, Li X. Nitrate pollution of groundwater in northern China. Agriculture, Ecosystems & Environment. 1996;59(3):223–31.
- 7. Hoffmann CC, Kronvang B, Audet J. Evaluation of nutrient retention in four restored Danish riparian wetlands. Hydrobiologia. 2011;674(1):5–24.
- 8. Ji G, Sun T, Zhou Q, Sui X, Chang S, Li P. Constructed subsurface flow wetland for treating heavy oil-produced water of the Liaohe Oilfield in China. Ecological Engineering. 2002;18(4):459–65.
- 9. Lind LP, Audet J, Tonderski K, Hoffmann CC. Nitrate removal capacity and nitrous oxide production in soil profiles of nitrogen loaded riparian wetlands inferred by laboratory microcosms. Soil Biology and Biochemistry. 2013;60:156–64.
- 10.
Buss S, Rivett M, Morgan P, Bemment C. Attenuation of nitrate in the subsurface environment. Environment Agency Science Group Report, SC030155/2. 2005.
- 11. Carrey R, Otero N, Soler A, Gómez-Alday J, Ayora C. The role of Lower Cretaceous sediments in groundwater nitrate attenuation in central Spain: Column experiments. Applied geochemistry. 2013;32:142–52.
- 12. Saunders D, Kalff J. Nitrogen retention in wetlands, lakes and rivers. Hydrobiologia. 2001;443(1):205–12.
- 13. Jain AK, Briegleb BP, Minschwaner K, Wuebbles DJ. Radiative forcings and global warming potentials of 39 greenhouse gases. Journal of Geophysical Research: Atmospheres. 2000;105(D16):20773–90.
- 14. Morse JL, Ardón M, Bernhardt ES. Using environmental variables and soil processes to forecast denitrification potential and nitrous oxide fluxes in coastal plain wetlands across different land uses. Journal of Geophysical Research: Biogeosciences. 2012;117(G2).
- 15. Hefting MM, Bobbink R, de Caluwe H. Nitrous oxide emission and denitrification in chronically nitrate-loaded riparian buffer zones. Journal of Environmental Quality. 2003;32(4):1194–203. pmid:12931872
- 16. Hayakawa A, Nakata M, Jiang R, Kuramochi K, Hatano R. Spatial variation of denitrification potential of grassland, windbreak forest, and riparian forest soils in an agricultural catchment in eastern Hokkaido, Japan. Ecological Engineering. 2012;47:92–100.
- 17. Zou Y, Duan X, Xue Z, Mingju E, Sun M, Lu X, et al. Water use conflict between wetland and agriculture. Journal of environmental management. 2018;224:140–6. pmid:30036808
- 18. Zhang L, Song L, Zhang L, Shao H, Chen X, Yan K. Seasonal dynamics in nitrous oxide emissions under different types of vegetation in saline-alkaline soils of the Yellow River Delta, China and implications for eco-restoring coastal wetland. Ecological engineering. 2013;61:82–9.
- 19. Song C, Zhang J, Wang Y, Wang Y, Zhao Z. Emission of CO2, CH4 and N2O from freshwater marsh in northeast of China. Journal of Environmental Management. 2008;88(3):428–36. pmid:17517465
- 20.
Federation WE, Association APH. Standard methods for the examination of water and wastewater. American Public Health Association (APHA): Washington, DC, USA. 2005.
- 21. Weatherburn M. Phenol-hypochlorite reaction for determination of ammonia. Analytical chemistry. 1967;39(8):971–4.
- 22. Burgin AJ, Hamilton SK, Gardner WS, McCarthy MJ. Nitrate reduction, denitrification, and dissimilatory nitrate reduction to ammonium in wetland sediments. Methods in Biogeochemistry of Wetlands. 2013;(methodsinbiogeo):519–37.
- 23.
Team RC. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. 2013. ISBN 3-900051-07-0; 2014.
- 24. Tongqian Z, Huashan X, Yuxiao H, Chao T, Hongqi M, Fanfu Z, et al. Agricultural non-point nitrogen pollution control function of different vegetation types in riparian wetlands: A case study in the Yellow River wetland in China. Journal of Environmental Sciences. 2009;21(7):933–9.
- 25. González-Alcaraz MN, Conesa HM, Álvarez-Rogel J. Nitrate removal from eutrophic wetlands polluted by metal-mine wastes: Effects of liming and plant growth. Journal of environmental management. 2013;128:964–72. pmid:23892281
- 26. Rivett MO, Buss SR, Morgan P, Smith JW, Bemment CD. Nitrate attenuation in groundwater: a review of biogeochemical controlling processes. Water research. 2008;42(16):4215–32. pmid:18721996
- 27. Brauer N, Maynard JJ, Dahlgren RA, O’Geen AT. Fate of nitrate in seepage from a restored wetland receiving agricultural tailwater. Ecological Engineering. 2015;81:207–17.
- 28. Mayer PM, Reynolds SK, McCutchen MD, Canfield TJ. Meta-analysis of nitrogen removal in riparian buffers. Journal of environmental quality. 2007;36(4):1172–80. pmid:17596626
- 29. Lin Y-F, Jing S-R, Lee D-Y, Chang Y-F, Shih K-C. Nitrate removal and denitrification affected by soil characteristics in nitrate treatment wetlands. Journal of Environmental Science and Health Part A. 2007;42(4):471–9.
- 30. Willems HP, Rotelli MD, Berry DF, Smith EP, Reneau Jr RB, Mostaghimi S. Nitrate removal in riparian wetland soils: effects of flow rate, temperature, nitrate concentration and soil depth. Water Research. 1997;31(4):841–9.
- 31. Dong LF, Sobey MN, Smith CJ, Rusmana I, Phillips W, Stott A, et al. Dissimilatory reduction of nitrate to ammonium, not denitrification or anammox, dominates benthic nitrate reduction in tropical estuaries. Limnology and Oceanography. 2011;56(1):279–91.
- 32. Gardner WS, McCarthy MJ. Nitrogen dynamics at the sediment–water interface in shallow, sub-tropical Florida Bay: why denitrification efficiency may decrease with increased eutrophication. Biogeochemistry. 2009;95(2–3):185–98.
- 33. Muñoz-Leoz B, Antigüedad I, Garbisu C, Ruiz-Romera E. Nitrogen transformations and greenhouse gas emissions from a riparian wetland soil: An undisturbed soil column study. Science of the Total Environment. 2011;409(4):763–70. pmid:21112616
- 34. Hoffmann CC, Rysgaard S, Berg P. Denitrification rates predicted by nitrogen-15 labeled nitrate microcosm studies, in situ measurements, and modeling. Journal of Environmental Quality. 2000;29(6):2020–8.
- 35. Stanford G, Vander Pol RA, Dzienia S. Denitrification Rates in Relation to Total and Extractable Soil Carbon 1. Soil Science Society of America Journal. 1975;39(2):284–9.
- 36. Puckett LJ, Zamora C, Essaid H, Wilson JT, Johnson HM, Brayton MJ, et al. Transport and Fate of Nitrate at the Ground-Water/Surface-Water Interface All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Journal of environmental quality. 2008;37(3):1034–50. pmid:18453426
- 37. Berg B. Litter decomposition and organic matter turnover in northern forest soils. Forest ecology and Management. 2000;133(1–2):13–22.
- 38. Esperschütz J, Zimmermann C, Dümig A, Welzl G, Buegger F, Elmer M, et al. Dynamics of microbial communities during decomposition of litter from pioneering plants in initial soil ecosystems. Biogeosciences. 2013;10(7):5115–24.
- 39. ŠImek M, Cooper J. The influence of soil pH on denitrification: progress towards the understanding of this interaction over the last 50 years. European Journal of Soil Science. 2002;53(3):345–54.
- 40. Giles ME, Morley NJ, Baggs EM, Daniell TJ. Soil nitrate reducing processes–drivers, mechanisms for spatial variation, and significance for nitrous oxide production. Frontiers in microbiology. 2012;3:407. pmid:23264770
- 41. Zaman M, Nguyen M, Saggar S. N2O and N2 emissions from pasture and wetland soils with and without amendments of nitrate, lime and zeolite under laboratory condition. Soil Research. 2008;46(7):526–34.
- 42. Wijler J, Delwiche C. Investigations on the denitrifying process in soil. Plant and Soil. 1954;5(2):155–69.
- 43. Koskinen W, Keeney D. Effect of pH on the Rate of Gaseous Products of Denitrification in a Silt Loam Soil1. Soil Science Society of America Journal. 1982;46(6):1165–7.
- 44. Jacinthe P, Bills J, Tedesco L, Barr R. Nitrous oxide emission from riparian buffers in relation to vegetation and flood frequency. Journal of Environmental Quality. 2012;41(1):95–105. pmid:22218178
- 45. Audet J, Hoffmann CC, Andersen PM, Baattrup-Pedersen A, Johansen JR, Larsen SE, et al. Nitrous oxide fluxes in undisturbed riparian wetlands located in agricultural catchments: Emission, uptake and controlling factors. Soil Biology and Biochemistry. 2014;68:291–9.
- 46. Hefting MM, Bobbink R, Janssens MP. Spatial variation in denitrification and N2O emission in relation to nitrate removal efficiency in a N-stressed riparian buffer zone. Ecosystems. 2006;9(4):550–63.
- 47. Hendzel L, Matthews C, Venkiteswaran J, St. Louis V, Burton D, Joyce E, et al. Nitrous oxide fluxes in three experimental boreal forest reservoirs. Environmental science & technology. 2005;39(12):4353–60.
- 48. Wu D, Dong W, Oenema O, Wang Y, Trebs I, Hu C. N2O consumption by low-nitrogen soil and its regulation by water and oxygen. Soil Biology and Biochemistry. 2013;60:165–72.
- 49. Chapuis‐Lardy L, Wrage N, Metay A, CHOTTE JL, Bernoux M. Soils, a sink for N2O? A review. Global Change Biology. 2007;13(1):1–17.
- 50. Blackmer A, Bremner J. Inhibitory effect of nitrate on reduction of N2O to N2 by soil microorganisms. Soil Biology and Biochemistry. 1978;10(3):187–91.
- 51. Weslien P, Kasimir Klemedtsson Å, Börjesson G, Klemedtsson L. Strong pH influence on N2O and CH4 fluxes from forested organic soils. European Journal of Soil Science. 2009;60(3):311–20.
- 52. Skiba U, Smith K. Nitrification and denitrification as sources of nitric oxide and nitrous oxide in a sandy loam soil. Soil Biology and Biochemistry. 1993;25(11):1527–36.
- 53. Van den Heuvel R, Bakker S, Jetten M, Hefting M. Decreased N2O reduction by low soil pH causes high N2O emissions in a riparian ecosystem. Geobiology. 2011;9(3):294–300. pmid:21504539
- 54. Butterbach-Bahl K, Baggs EM, Dannenmann M, Kiese R, Zechmeister-Boltenstern S. Nitrous oxide emissions from soils: how well do we understand the processes and their controls? Phil Trans R Soc B. 2013;368(1621):20130122. pmid:23713120