Next Article in Journal
A Two-Stage Model for Data-Driven Leakage Detection and Localization in Water Distribution Networks
Previous Article in Journal
Three-Dimensional Smooth Particle Hydrodynamics Modelling of Liquid–Sediment Interaction at Coastline Region
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Water
Volume 15
Issue 15
10.3390/w15152709
water-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Measuring Nitrate Leaching in the Vadose Zone of Loess Soils—Comparison of Batch Extraction and Centrifugation

by
Dico Fraters
1,*,
Gerard H. Ros
2,3 and
Timo Brussée
1,*
1
National Institute for Public Health and the Environment (RIVM), P.O. Box 1, 3720 BA Bilthoven, The Netherlands
2
Nutrient Management Institute BV, Nieuwe Kanaal 7c, 6709 PA Wageningen, The Netherlands
3
Environmental Systems Analysis Group, Wageningen University & Research, P.O. Box 47, 6700 AA Wageningen, The Netherlands
*
Authors to whom correspondence should be addressed.
Water 2023, 15(15), 2709; https://doi.org/10.3390/w15152709
Submission received: 26 June 2023 / Revised: 20 July 2023 / Accepted: 25 July 2023 / Published: 27 July 2023
Browse Figures
Review Reports Versions Notes

Abstract

:
The nitrate concentration in the subsoil moisture of the vadose zone is an important indicator for future groundwater quality, which is classically determined via centrifugation. Batch extraction is an inexpensive and easy alternative method, but whether these methods measure the same soil water, nitrogen species, and nitrate concentrations is unclear, in particular for loess soils. Two experiments were carried out to assess the differences in nitrate and other anion concentrations between centrifugated soil moisture (centrifugated at different speeds and times) and batch extractions (using double-distilled water and 0.01 M CaCl2). Batch extraction resulted in lower nitrate (−20%) and chloride (−15%) concentrations than centrifugation, mainly due to anion exclusion, where soil microporosity controls the contribution of diffusion, denitrification, and leaching processes. Vice versa, batch extraction overestimated the concentration of nutrients that occur as precipitates in or sorb the soil matrix, such as sulphate (+50%) and ammonium (+96%). Batch extractions can only be used as a proxy to determine actual nitrate concentrations of soil water. However, they are useful to monitor changes in nitrate leaching over time in response to (policy) measures taken. They can also be used as “early warning indicator” and to improve the reliability of spatial explicit monitoring networks.

1. Introduction

Nitrogen (N) is an important nutrient in agriculture. Since the 1940s, European agriculture has intensified, greatly relying on the use of large amounts of N fertilisers and manure to accelerate crop production. A substantial part of the added nitrogen is lost to the environment, since only 60 to 65% of the added N is taken up by crops in the European Union [1]. The N surplus, being the N not taken up by crops, is largely lost to the environment, with undesired impacts on air, groundwater, and surface water quality. This poses a threat to both the aquatic environment and human health [2,3]. Policies have, therefore, been adopted to reduce the negative impacts of N use, including the National Emission reduction Commitments Directive (for ammonia and NOx), the Birds and Habitats Directive (for biodiversity), the Water Framework and Nitrates Directive (for water quality), and the Paris Climate Agreement [4] (for greenhouse gases). Through the so-called Farm to Fork Strategy, the EU Member States aim to establish a sustainable food production system characterised by high nutrient use efficiencies to protect the environment, preserve biodiversity, and tackle climate change [5]. Agriculture has, therefore, to increase the nitrogen use efficiency (NUE) by optimising crop, soil, and fertiliser management.
The Nitrates Directive additionally obliges Member States to implement measures to decrease the N losses to the groundwater and surface water and to monitor the current state, as well as the impact of measures taken to reduce the losses. In the Netherlands, the nitrate concentration in water leaching from the root zone of agricultural fields is, therefore, monitored at 450 farms, representing more than 80% of the agricultural land [6]. Nitrate concentrations are measured in pore water being extracted via centrifugation.
There are several methods to sample root zone leachate within the vadose zone, which differ in complexity, cost, and labour intensity [7,8,9,10,11,12,13,14,15]. Centrifugation [16,17], pressure filtering or squeezing [18,19], and batch extraction [20,21] are often used for the soil moisture extraction from soil samples taken from the soil layer directly below the rooting zone. Centrifugation has been shown to sample the same water pool (low tension, mobile water) isotopically as (non-destructive) suction cups [22], whereas methods using vapour equilibration, microwave extraction, or cryogenic extraction might affect the composition of the water extracted [23]. Centrifugation is often usen in monitoring programs as a robust, reproducible, and standardised way of determining solute concentrations in pore water in silt loam subsoils [24,25], including the assessment of ion concentrations in pore water [26,27] and the monitoring of nitrate leaching [28,29]. However, centrifugation is more costly and labour-intensive than batch extraction, where the latter might offer an alternative and less expensive method to facilitate national monitoring of nitrate leaching.
Batch extraction is a simple extraction method [30]. A soil sample is split into two subsamples. A certain amount of extraction solution is added to the first subsample. Once the solution is in equilibrium with the soil moisture, the solution and soil are separated, and the concentration in the solution is determined. The other subsample is used to determine the soil moisture content and, subsequently, a dilution ratio based on both subsamples. The concentration in the soil moisture of the undisturbed soil sample is calculated using the dilution ratio given the moisture content of the soil. Batch extractions are commonly used for routine agricultural soil analyses, as well as environmental assessments [30,31,32], but it remains unclear whether both methods extract the same water pool and are able to detect (changes in) nitrate concentrations in the vadose zone of loess soils. This is particularly true for nitrogen species since the extraction conditions might impact both the N species determined and their concentration [20,21,33], in particular for ammonium and extractable organic nitrogen [34].
Variation in the extraction conditions leads to differences in the occurrence of geochemical processes in the soil sample and, thus, affects the measured ion speciation and concentration in the soil moisture. Anions such as nitrate and chloride, however, are assumed to be hardly influenced by the adsorption–desorption or precipitation–dissolution processes, specifically in soils without anion exchange capacity [35]. This suggests that both batch extraction and centrifugation can be used to measure nitrate concentrations in the vadose zone of agricultural soils. However, the transport of anions in loamy and clayey soils is known to be influenced by anion exclusion [36]. Due to the negatively charged particles in the walls of pores, anion concentrations in soil moisture are lower near the wall and higher near the centre of the pore [37]. Although the effect of anion exclusion on transport has been studied extensively [38,39], the implications for the sampling methodology to assess the risk of nitrate leaching remains unknown. For example, Reference [40] concluded that dilution effects for nitrate were adequately accounted for by assuming no buffering by soil particles for three clay loams with strongly different parent materials (basalt, sandstone, and siltstone). Similarly, Reference [20] reported no consistent impacts of the soil-to-solution ratio on nitrate concentrations in loess topsoil samples. Nevertheless, Reference [41] reported an increase in nitrate concentrations when they decreased the soil-to-solution ratio using silty and clay-rich subsoil samples. Furthermore, Reference [42] showed that nitrate is partly immobile in loess subsoil and that, by choosing a particular extraction solution, one can select the nitrate compounds that are sensitive to leaching.
There have only been a few studies that have investigated both the batch extraction and centrifugation methods [19,43,44,45,46,47] while focussing on the differences for dissolved organic nitrogen compounds, electric conductivity, and inorganic contaminants. As far as we know, there are no studies that compare batch extraction and centrifugation with regards to the nitrate concentration in soil moisture. This raises the question of whether batch extraction can be used instead of centrifugation to determine the nitrate concentrations in soil moisture in the unsaturated subsoil and, secondly, what the potential consequences are for the estimated nitrate leaching risks to groundwater.
The objective of this study was, therefore, to determine whether batch extraction could be used instead of the centrifugation method to measure the nitrate concentration in soil moisture within the unsaturated zone of loess soils. This study focused on two regularly used batch extraction methods using fresh soil samples, which were extraction with double-distilled water and extraction with a 0.01 M CaCl2 solution. We compared the nutrient concentrations between centrifugation and batch extraction, as well as their ratios (NO3 to Cl and SO4 to Cl), assessed the potential impact of anion exclusion, and explored whether differences between both methods can be quantified in response to a variation in the soil properties.

2. Materials and Methods

2.1. General Setup of the Study

Two experiments were designed to unravel differences between nitrate concentrations in soil moisture determined by batch extraction and by drainage centrifugation. The first experiment aimed to determine how much residual nitrate remains after drainage centrifugation (accounting for differences in centrifugation conditions) and can be extracted by batch extraction afterwards (details in Section 2.4). The second experiment compared batch extraction and centrifugation for a series of soils (n = 14 soils, 3 depths, 2 duplicates), where both methods were used (details in Section 2.5). The batch extractions included extractions with pure water and 0.01 M CaCl2, both performed at relatively small soil-to-solution ratios (1:1 to 1:2 v/v). Differences between pure water and CaCl2 effects clarified the impact of anion exclusion, where pure water would lower the ionic strength and, thus, lead to another equilibrium for substances that may be adsorbed by the soil matrix or can precipitate or dissolve. The relatively small soil-to-solution ratios were selected to maximise the potential differences between both methods: (i) the smaller the ratio, the smaller the effect on the calculated concentration in the soil solution is; (ii) it minimises the number of values below the limit of detection.

2.2. Site Description and Sampling

Soil was sampled from agricultural fields in the loess region in the southern part of the Netherlands (see Figure 1). The landscape is slightly undulating (20–325 m above sea level). In the western part of the region is the river plain of the River Meuse, which flows south to north in this region. To the east of this river plain are plateaus covered with a 2–20 m-thick layer of loess [48]. The main soil types are characterised as Eutric-Gleyic Fluvisols in the river plain and Haplic Luvisols outside the river plain [49]. The climate is temperate maritime, with annual precipitation ranging from 775 mm in the northwestern part to 925 mm in the southeast [50].
Soil samples for Experiment 1 were collected in 2014 at four fields (A to D, Figure 1), following standardised procedures described by [24]. At each location, five drillings were carried out (semi-replicates). Each drilling resulted in eight replicate samples of the subsoil, representative of the subsoil between 1.5 and 3.0 m below the surface. In total, 40 (semi-)replicate samples of 0.5 kg were collected for each of the four locations.
Soil samples for Experiment 2 were collected in 2015 from 14 fields within drinking water protection areas (light blue areas, Figure 1). In each of the selected fields, a borehole was drilled using an Edelman auger at two locations, representing different conditions within that field. Per borehole, duplicate samples were taken at three depths (3.3–3.5, 4.3–4.5, and 5.3–5.5 m below the surface), unless sampling was restricted by the presence of chalk or gravel. At each sampling depth, two samples were taken using stainless steel tubes (length 0.240 m, diameter 0.038 m, content 0.226 L). The tubes were sealed and stored in a dark and cool environment until processing. In total, 77 samples were collected.

2.3. Soil Analysis

Standard soil analyses were performed for all samples of Experiment 1 by the Eurofins Agro laboratory in accordance with Dutch or international standard procedures: soil texture (NEN 5753; sand (>50 μm) gravimetric, clay (<2 μm) density measurement using Stokes formula, silt calculated), organic matter (NEN 5754, weight loss at 550 °C), inorganic carbon (ISO 10694, 1000 °C with infra-red spectrophotometry), cation exchange capacity (CEC) (ISO 23470, extraction with 0.0166 M cobalt hexamine tri-chloride), and pH (NEN-ISO 10390, 1:10 v/v 0.1 M KCL and 1:10 v/v water suspension). Soil moisture content was determined for a separate replicate sample per borehole by weight loss at 105 °C for 48 h in the laboratory by TNO.

2.4. Experiment 1: Centrifugation and 1:1 Water Extraction

Soil samples were removed from the refrigerator and stored at room temperature in the dark for at least 12 h before processing to avoid condensation on the wall of the container. Replicate samples were centrifuged at five centrifugation time levels and five relative centrifugal force levels to elucidate the impact of both factors on the ion composition and ion concentration of the centrifuge. A preliminary test was performed with the soils of Location D to quantify the combined effect of force and time on soil moisture extraction efficiency. The soils from Locations A to C were centrifuged at two force levels (733 g and 6579 g) and five time levels ranging from 20 to 240 min in order to quantify the impact of centrifugation time. In addition, soils from Locations A to D were centrifuged at one time level (35 min) and five force levels ranging from 117 to 14,191 g. In total, 48 solute concentrations were available for the data analyses (time series: 2 force levels × 5 time levels × 3 locations = 30, force series: 5 force levels × 1 time level × 4 locations = 20). Further details about the experimental setup and chemical analysis are given in [24].
After centrifugation, the samples were stored in glass containers in a refrigerator until processing for batch extraction. When processed, each centrifuged soil sample was mixed briefly (30 s) in a 2 L polypropylene box with a porcelain spatula and subsampled for the determination of the remaining moisture content and batch extraction. Batch extractions were carried out in triplicate (see Figure 2). In summary, 50 g of soil was placed in a 500 mL polyethylene jar, and then, 50 g of Milli-Q water was added. The jars were closed with a lid and placed on a horizontal shaker (IKA HS501 digital) during 60 min (200 min−1 at 30 mm). The solution was separated from the soil by centrifugation in glass centrifugation tubes (Sorvall T6000 D, 60 min at 1900 rpm) and then filtered over a 0.45 μm polyethersulfone filter using a polypropylene syringe. Soil moisture was determined by drying 100 g of soil for 24 h at 105 °C. Solution samples were analysed for nitrate, chloride, and sulphate via ion chromatography (internal procedure AC-W-066) and for ammonium (acidification with H2SO4, pH < 2; spectrophotometry-continuous flow analysis; internal procedure AC-W-027).

2.5. Experiment 2: Centrifugation and 1:2 CaCl2 Extraction

Soil samples were split, where one part was centrifuged at 6597 g for 35 min (see Figure 2) in accordance with the procedure described by [24] and one part was extracted via batch extraction. Centrifugated solutions were analysed for nitrate, chloride, sulphate, and ammonium, as described above. The sample for batch extraction was subsampled for (i) the determination of moisture content by drying 50 mL of soil at 105 °C until its weight did not change and for (ii) batch extraction. For batch extraction, soil was added to a glass vessel with 0.100 L of 0.01 M CaCl2 solution up to a volume of 0.150 L (soil-to-liquid ratio 1:2 v/v) in accordance with the method of [51]. The vessels were placed on a horizontal shaker for 60 min (200 min−1 at 30 mm). The solution was separated from the soil by filtration using a 0.45 μm filter. The solution was analysed for nitrate and ammonium by means of colorimetry [51].

2.6. Data Analysis

For batch extraction, the solute concentration in the soil moisture (C0) was calculated by multiplying the measured concentration in the batch extract (Cb) by a dilution ratio. This dilution ratio was calculated based on the soil moisture content (W) measured in the first subsample, the weight of the solution added (Vsolution) to the second subsample used for extraction, and the weight of this fresh (wet) sample (Vsoil):
C0 = Cb × (Vsolution × (1 + 1/W) + Vsoil)/Vsoil
If the weight of the added solution and the fresh soil sample were exactly the same, this formula was reduced to: (1)
C0 = Cb × (2 + 1/W)
For Experiment 1, the soil moisture solute concentration by batch extraction of a fresh (non-centrifuged) soil sample was calculated as the weighted mean of the concentration in the centrifugation extract and the concentration in the batch extract of the already-centrifuged soil sample.
Data controls discovered unrealistic results in Experiment 1 for one of the triplicate batch samples from Location B, which was centrifuged at 7500 rpm for 35 min before batch extraction. Nitrate, chloride, or sulphate concentrations in this sample were all below the limit of detection, while in the other two triplicates, the concentrations found were 186 mg/L, 35 mg/L, and 66 mg/L, respectively. Therefore, this single sample was removed before analysis. All nitrate data were used in the analyses of Experiment 2, using the detection limit values when concentrations were not detectable, but additional analyses were performed without values below the detection limit to check their effect on the results.
Data handling and statistical analysis were performed with R (4.2.1) [52]. For data handling, the tidyverse package was used [53]. As multiple samples were collected in each of the fields for both experiments, mixed-effect models were used to analyse the difference in nitrate, ammonium, chloride, and sulphate concentration between the batch and centrifugation method, their relationship, and the effect of other parameters on the difference between the methods. For this modelling, the nlme package [54] was used with “field” as the random variable in the first experiment and “sampling point within field” in the second experiment. The confidence interval (CI) and prediction interval (PI) were calculated as described in [55]. To estimate the marginal means, the packages emmeans was used [56].

3. Results

3.1. Experiment 1

Soil moisture content ranged from 17.2 to 25.7% (w/w) with a mean of 21.9%. The concentrations of nitrate, ammonium, chloride, and sulphate in the centrifugation and batch extracts are given in Table 1 and shown in Figure 3. The ammonium was below the limit of detection in only 3% of the centrifugation extracts (LoD 0.064 mg/L), but in all the batch extracts (LoD 0.49 mg/L). Concentrations of the other solutes were always above the LoD. The nitrate and chloride concentrations in the centrifugation extracts were higher than those in the batch extracts, on average, 31% and 43%, respectively (Table 1). The soil moisture recovery (SMR), i.e., the percentage of a sample’s total soil moisture extracted by centrifugation, ranged from 4.1 to 39% with a mean of 22%. This means that 27–28% of the total extractable nitrate and chloride was extracted by centrifugation. The ammonium concentration was substantially lower than the nitrate concentration, and it was not meaningful to compare centrifugation and batch extraction, as the batch concentrations were below the LoD (Table 1 and Figure 3B). The sulphate concentration in the centrifugation extract was 26% lower than in the batch extract (Figure 3D). This is in contrast to the two other anions. The recovery of sulphate from the sample with centrifugation was, therefore, lower than for nitrate and chloride.
Table
Table 1. Comparison of mean concentrations of nitrate, ammonium, chloride, and sulphate (mg/L) in soil moisture determined by centrifugation and 1:1 (w/w) Milli-Q water batch extraction, the mean difference between both, and the associated t-value.
Table 1. Comparison of mean concentrations of nitrate, ammonium, chloride, and sulphate (mg/L) in soil moisture determined by centrifugation and 1:1 (w/w) Milli-Q water batch extraction, the mean difference between both, and the associated t-value.
ParameterCentrifugation 1Batch 1:1 2Batch 1:1 3Difference 4t-Value 4
Nitrate153 (8.7–346)108116 (6.5–234)343.03
Ammonium 50.17 (<0.064–0.45)0.200.18 (<0.073–0.35)−0.01−0.49
Chloride24.5 (11.9–50.7)16.117.7 (5.4–40.6)6.55.47
Sulphate29.4 (16.2–55.3)42.740.4 (21.0–66.2)−10.3−3.55
Note(s): 1 Concentration in centrifugation extract; mean, with minimum and maximum between brackets. 2 Concentration in batch extract of sample already centrifuged; corrected for dilution. 3 Calculated weighed mean concentration of centrifugation and batch results as estimated for fresh soil sample batch extraction; mean, with minimum and maximum between brackets. 4 Mean and t-value of difference between individual centrifugation (1) and recalculated batch (2) results. 5 Ammonium concentrations in batch extracts were all below the limit of detection (LoD); 3% were below in centrifugation extracts.
Water 15 02709 g003 550
Figure 3. Relationship between nitrate (A), ammonium (B), chloride (C), and sulphate (D) concentration in soil moisture determined by a 1:1 Milli-Q water batch extraction (x-axis) and by centrifugation (y-axis) using samples from Locations A to D. The intercepts and regression coefficients and their significance are given in Table 2. The limit of detection (LoD) is shown for ammonium (red dotted line). Confidence (Conf.Int) and prediction (Pred.Int) intervals are shown in dark and light pink, respectively.
Figure 3. Relationship between nitrate (A), ammonium (B), chloride (C), and sulphate (D) concentration in soil moisture determined by a 1:1 Milli-Q water batch extraction (x-axis) and by centrifugation (y-axis) using samples from Locations A to D. The intercepts and regression coefficients and their significance are given in Table 2. The limit of detection (LoD) is shown for ammonium (red dotted line). Confidence (Conf.Int) and prediction (Pred.Int) intervals are shown in dark and light pink, respectively.
Water 15 02709 g003
Table
Table 2. Relationship between solute concentration in centrifugation (dependent) and batch (independent) extracts for nitrate, ammonium, chloride, and sulphate, based on mixed model analyses (intercept and regression coefficient (mean), their standard error (SE), and p-value).
Table 2. Relationship between solute concentration in centrifugation (dependent) and batch (independent) extracts for nitrate, ammonium, chloride, and sulphate, based on mixed model analyses (intercept and regression coefficient (mean), their standard error (SE), and p-value).
ParameterExperimentIntercept Coefficient
MeanSEp-Value 1MeanSEp-Value 1
nitrate11.815.370.731.270.038<0.001
nitrate217.98.800.0461.090.0910.33
ammonium10.0770.0650.240.510.3600.18
chloride13.460.33<0.0011.170.014<0.001
sulphate13.076.060.610.660.1280.012
Note: 1 p-value based on H0: intercept = 0 and coefficient = 1.
The NO3-to-Cl ratio in the centrifugation extracts was about 87% (p < 0.001) of the ratio in the water batch extracts (Figure 4A), showing that relatively less nitrate was extracted with centrifugation than with batch extraction. This difference was even bigger for sulphate, where the SO4-to-Cl ratio in the centrifugation extracts was about 45% (p < 0.001) of the SO4-to-Cl ratio in the batch extracts (Figure 4B).
An increase in SMR led to a smaller difference in the chloride concentration (p < 0.001) between the two methods (Figure 5A) and to a larger (more negative) difference in the sulphate concentration (p < 0.001, Figure 5B). Nitrate showed a similar, but statistically insignificant (p = 0.28), behaviour as chloride. The difference in chloride concentration between centrifugation and batch extracts decreased when the CEC of the soil sample was larger (p < 0.001, Figure 6).

3.2. Experiment 2

Soil moisture content ranged from 16.6 to 23.8% (w/w) in the duplicate samples used for centrifugation and ranged from 16.5 to 27.8% (w/w) in the duplicate samples used for batch extraction. The mean moisture content was 17.2% in both duplicate sets. The centrifugation SMR ranged from 10 to 41% with a mean of 30%. The nitrate concentration ranged from 0.21 to 476 mg/L with a mean of 92 mg/L in centrifugation extracts and from below the detection limit to 450 mg/L with a mean of 68 mg/L in batch extracts (Figure 7). Centrifugation extracted about 40% of the total extractable nitrate. About 30% of the nitrate concentrations in batch extracts were below the limit of detection (0.6 mg/L as NO3-N). Accounting for the dilution of the soil solution by the addition of CaCl2, this means a detection limit between 8 and 16 mg/L as NO3 in the soil moisture. The ammonium concentrations in batch extracts were all below the limit of detection (0.5 mg/L as NH4-N). In the centrifugation extract, 30% of the ammonium concentrations was below the detection limit of 0.05 mg/L, with a mean concentration of 0.12 mg/L and a range of <0.05 to 1.02 mg/L. The chloride concentration in the centrifugation extract ranged from 1.5 to 80 mg/L with a mean of 27.5 mg/L and the sulphate concentration from 1.7 to 447 mg/L with a mean of 55 mg/L.
The nitrate concentration (p = 0.68) and the soil moisture content (p = 0.126) did not change with depth. The chloride concentration increased with depth (p = 0.016, 4 mg/L/m), while the sulphate concentration decreased (p = 0.0175; 17 mg/L/m). Nevertheless, the nitrate-to-chloride and nitrate-to-sulphate ratios did not change significantly with depth, at p = 0.251 and p = 0.149, respectively. The nitrate concentration was on average 36% higher in the centrifugation extract than in the CaCl2 batch extract (Table 3; Figure 7). Similar findings were found when the nitrate concentrations below LoD were removed prior to the analysis.
The relative difference in nitrate concentration between the two methods, expressed as the relative difference compared to the batch concentration, was not influenced by the nitrate concentration in the batch extract (p = 0.32; Figure 8A), although substantial deviations occurred across the range of nitrate concentrations. The difference was also not related to the depth of sampling (p = 0.44; Figure 8B), the soil moisture content (p = 0.36; Figure 8C), or the SMR (p = 0.88; Figure 8D).

4. Discussion

The objective of this study was to determine whether batch extractions can replace drainage centrifugation to measure the nitrate concentration in soil moisture within the unsaturated zone of loess soils. There was sufficient evidence that the nitrate concentration differed between both methods. We first discuss these differences in relation to spatial microvariation within the vadose zone. Thereafter, we look into the impact of the extraction methodology and sample type and whether this can help to explain the differences we and others have found. We conclude with a discussion of the opportunities to estimate nitrate leaching to groundwater based on batch extraction.
  • Spatial microvariation in nitrate concentration
The extraction method to collect soil moisture from the vadose zone and the associated nitrate analysis had a substantial impact on the nitrate concentration determined in the soil moisture that leaches to the groundwater. Both experiments showed that the nitrate concentration was 30–40% higher in the centrifugation extract compared to the batch extract. This contradicts the classic hypothesis that nitrate is not affected by adsorption–desorption or precipitation–dissolution in soils without anion exchange capacity. The variation of nitrate concentrations in pore water showed that the concentration was not homogeneous and calls into question whether batch extraction can be used to provide an unbiased estimate of the nitrate concentration in pore water leaching to groundwater. In contrast to our results, Reference [57] reported a higher nitrate concentration in a batch extract of a shallow sample of a sandy loam (0.2–0.3 m depths) than in a porous cup extract, but—similarly to our results—the concentration in the batch extract of subsoil samples (0.7–0.8 m) was lower than in the porous cup extract. Porous cups, just like centrifugation, only extract part of the soil moisture. Reference [57] attributed this different behaviour between topsoil and subsoil to the production of nitrate (nitrification) in soil aggregates occurring mainly in the topsoil, suggesting that the distribution of nitrate in the soil moisture changes across the soil profile.
The fact that the difference between the nitrate concentration in centrifugation- and batch-extracted soil moisture is not the same for topsoil and subsoil samples might also be due to a variation in water potential within the soil matrix. In the topsoil of most agricultural fields, the nitrate is predominantly available in the immobile water located within the soil aggregates, as suggested by [35]. As percolation occurs, porous cups sample the mobile water with lower nitrate concentrations. At greater depths, the water leaching from above contains a higher concentration than the immobile water in subsoil aggregates. Consequently, centrifugation and lysimetry would access mostly continuous water, while batch extraction would include isolated water as well. This isolated water is present in areas in the soil matrix where water movement is restricted by the physical structure of the soil. As such, we would expect that differences between both methods decline with a decrease in soil clay content, a finding confirmed by others [57,58]. Both studies found no differences in the nitrate concentration between batch extraction and porous cups in sandy topsoils. In addition, Reference [59] showed that, in the upper layer of a loamy soil (0–0.15 m), the nitrate concentration was also higher in a centrifugation extract than in soil moisture sampled by porous cups and zero-tension lysimeters, whereas the nitrate concentrations did not differ in the deeper soil layer (0.30–0.60 m). Again, soil microporosity seemed to control the nitrate concentration in the topsoil. We showed that, in centrifugation extraction, only part of the soil moisture (i.e., only the mobile water in the large pores) is extracted, whereas in batch extraction, all the soil moisture (i.e., also the immobile water in small pores) is extracted. Given the higher concentration in the soil moisture extracted by centrifugation, we concluded that the immobile water in the subsoil has a lower nitrate concentration than the mobile water.
A lower nitrate concentration in immobile water in subsoil micropores could be due to denitrification as a consequence of local anaerobic conditions within soil aggregates [60]. Denitrification is a common process controlled by carbon and nitrogen availability, temperature, and anaerobic conditions. Reference [61] did indeed show that long-term (30-year) nitrogen fertilisation increased the abundance of denitrifiers in loess subsoils. In aggregated soil, however, it is important to consider diffusion to the centre of aggregates, where denitrification predominately occurs [62]. For well-aerated soil with air-filled macropores, the critical distance for anoxic conditions in aggregate micropores might not be exceeded [63]. Denitrification in micropores might have caused a decline in nitrate concentrations in our experiments, but since chloride showed a similar behaviour as nitrate, there needs to be another process controlling the nitrate concentrations in immobile water. The most-likely process to explain the observed differences in both the chloride and nitrate concentration is anion exclusion, as this process leads to a higher anion concentrations in the more mobile water than in the less-mobile water in the unsaturated zone [37,64]. Anion exclusion is the process by which anions, such as nitrate, are repelled by the negatively charged clay and organic matter particles in a thin layer of water surrounding soil particles in pores. As a result, the nitrate concentration is lower in the thin layer of water around soil particles than in the surrounding water.
We would, therefore, expect that a higher CEC leads to a more-pronounced anion exclusion and, therefore, a larger difference between the chloride concentration in the centrifugation and batch extracts. However, the opposite is the case (Figure 6). The underlying processes remain unclear. Though centrifugation extracts predominantly the mobile water, part of the less-mobile water can be extracted when the conditions during centrifugation are more severe. As [59] showed, centrifugation makes it possible to extract part of the less-mobile water. In agreement with the finding of [59], an increase in SMR, i.e., a larger contribution of relatively less-mobile water in the centrifugation extract, resulted in a smaller difference in the chloride concentration between centrifugation and batch extracts in Experiment 1 (Figure 5A).
As stated, both nitrate and chloride had a higher concentration in centrifugation extracts than in batch extracts. In Experiment 1, the same percentage of the total amount was extracted by centrifugation for both solutes (27–28%), and both solutes exhibited a similar relationship between concentrations in the centrifugation and batch extracts (Figure 3A,C). In Experiment 2, the nitrate-to-chloride ratio in the centrifugation extract did not significantly change with sampling depth (mean ratio of 2.3, trend −0.2 m−1; p = 0.251; see Appendix A), which could be an indication that denitrification did not have a large impact on the nitrate concentration in the mobile pore water. Therefore, we assumed that anion exclusion is the main cause of the difference in concentration between centrifugation and batch extraction.
However, if anion exclusion is the main factor causing a higher nitrate and chloride concentration in the centrifugation extract, we would expect an even stronger effect on the sulphate ion, as it has a more negative charge than chloride and nitrate [65]. However, the Experiment 1 results showed the opposite behaviour for sulphate (Figure 3D). This could be explained by precipitation, where the sulphate is dissolved again when the soil sample is mixed with water and shaken. High wet and dry sulphate deposition originating in the past might have resulted in stable sulphate minerals [66]. Reference [66] reported that the maximum sulphate content in the soil profile was reflected in the sulphate concentrations in soil moisture extracted by squeezing, but not in porous cup measurements. In addition, Reference [67] studied the origin of sulphate in the vadose zone and groundwater of a loess aquifer in China and concluded that water-extractable concentrations of sulphate are not representative of the actual sulphate concentration due to the presence of gypsum, either in closed pores or covered by a calcium carbonate film. Based on our findings, we postulated that this is also the case in Dutch loess soils.
  • Impact of extraction methodology and type of sample
The results of batch extractions are strongly influenced by soil pre-treatment [20,21,33], in particular when soils are dried and sieved. The use of pure water may change the pH of the soil moisture and, thereby, the chemical equilibrium. For that reason, References [21,51] proposed the use of a 0.01 M weak salt solution to mimic the actual conditions occurring in the soil moisture in agricultural soils. Increasing the ionic strength additionally affects the release of potential plant-available nutrients. Salt solutions that are often used to mimic the N concentrations in soil moisture include 0.01 M CaCl2 and 1 M and 2 M KCl. The extraction methodology will have a substantial effect on ammonia and dissolved organic nitrogen [34], but might also impact the nitrate. For example, Reference [68] showed that a batch extraction (soil-to-solution ratio of 1:10) with 0.01 M BaCl2 extracted more nitrate than 0.02 M SrCl2 and 0.01 M CaCl2 in acid and neutral soils (29–35% clay), but not for basic soils (5–21% clay). In addition, Reference [20] found inconsistent results when comparing 1 M KCl, 0.01 M CaCl2, and 0.5 M K2SO4 solutions (soil-to-solution ratio of 1:10), but increasing the molarity of the KCl solution from 0.5 to 1 M resulted in the extraction of a smaller amount of nitrate. There was no difference between the 1 M and 2 M solutions. However, these empirical studies did not provide any insight into the potential causes of the differences we found. Our results showed that centrifugation resulted in a higher nitrate concentration than both a 1:1 pure water (+31%) and a 1:2 0.01 M CaCl2 solution batch extraction (+36%) of silt loam subsoil samples. The results for both batch extraction methods were similar, even though both the ionic strength and the soil-to-solution ratio were different.
A difference in the soil-to-solution ratio in batch extraction will influence the release of nutrients for which the concentration is controlled by the adsorption–desorption and/or precipitation–dissolution processes [46], such as ammonium. Due to low ammonium concentrations and a relatively high LoD for batch extraction, the current study could not confirm this, even though the soil-to-solution ratios were high. For soils with anion exchange capacity (AEC), the soil solution nitrate concentration, corrected for dilution, is affected by the soil-to-solution ratio [69]. For anions such as nitrate and chloride, one would assume that the soil-to-solution ratio, the type of extraction solution, and the extraction time and condition have hardly any effect on the results of batch extraction in soils without AEC. Indeed, Reference [43] concluded that each of the three soil-to-water-extraction ratios tested (1:2, 1:5, and 1:10) were effective in estimating the electric conductivity of a saturated paste extract, but they recommended adopting a ratio of 1:2 to reduce possible issues with dilution. Other studies also showed no effect of the dilution ratio on the nitrate concentration when corrected for dilution. Reference [20] reported the highest nitrate extraction for a soil-to-solution ratio of 1:10, while both 1:5 and 1:20 ratios gave lower amounts, which seems to suggest that the ratio has no influence on the nitrate concentration. Nevertheless, Reference [41] reported an increase in extracted nitrate when they decreased the soil-to-solution ratio from 1:1 to 1:10. They used a 2 M KCl solution to extract silty to clay-rich subsoil samples from loess soils. The isotopic signature of NO3-N showed incomplete extraction at the 1:1 ratio. They assumed that incomplete dispersion of particles caused the lower nitrate extraction efficiency. Reference [42] showed that, for a subsoil without AEC, but rich in smectite (>40%), a 1 M KCl solution extracted more nitrate than double-distilled water (DDW). They postulated that the KCl solution collapsed the interparticle spaces and removed both bound and DDW-extractable nitrate. They concluded that nitrate concentrations obtained by DDW extraction best represent a nitrate-in-soil solution that would be vertically transported with recharge. However, these results cannot explain why the nitrate concentration in a centrifugation extract is higher than in a batch extract, as centrifugation leaves the soil structure more or less intact.
Finally, soil type and sampling depth may also contribute to differences between extraction methods given the variation in compaction, elemental composition, and exchange capacity. For example, Reference [40] showed that nitrate concentrations in paste extracts—when accounting for dilution—were in good agreement with nitrate concentrations in soil moisture extracted by porous cups. They compared the results of a paste batch extraction (240 g soil: 60 mL water) to those acquired by porous cup extraction and showed that nitrate concentrations at other soil-to-solution ratios (up to 1:16), when corrected for dilution, were not significantly different from the concentrations in the saturated paste. However, the authors studied the upper 0.1 m of clay loam soils with organic carbon contents ranging from 1.7–8,6%, while in our study, we used silt loam subsoil material (>1.5 m depths) with a low organic carbon content (0.01–0.72%). As discussed above, the extraction of samples from the topsoil might lead to other results than the extraction of samples from the subsoil. The findings of [70], however, seem to be in agreement with our findings. At one location—a sandy loam field—[70] nitrogen leaching was measured at a depth of about 1.3 m during 2 y. The authors compared differences between free-draining lysimeters, porous cups (10 kPa suction), and soil core batch extracts (2 M KCl, with a soil-to-solution ratio of 1:5) and found that the measured nitrate leaching with lysimeters was 12% higher than the leaching calculated using batch extracts. Calculations using porous cup measurements resulted in 21% higher values. These differences were explained by (a) lower nitrate concentrations in subsoil micropores, which were not sampled by lysimeters and porous cups, (b) nitrogen mineralisation in soil core samples due to freezing and thawing before extraction, and (c) a high limit of detection due to the dilution of the soil moisture. The latter two reasons could not have played a role in our experiments. Our samples were not stored frozen, and the detection limit was not as big of a problem, as we used a narrow soil-to-liquid ratio.
  • Estimating nitrate concentration leaching to groundwater
The results of this study imply that the nitrate and chloride concentrations were higher in the more-mobile soil moisture (up to pF 3.6) compared to the less-mobile moisture fraction (pF > 3.6) of loess soils. Earlier research showed that concentrations within the more-mobile fraction of soil moisture do not depend on the tension applied at centrifugation [24,58]. Whether or not batch extraction can be used instead of centrifugation to estimate the nitrate concentration in the actual nitrogen leaching from the root zone depends on the aim and setup of the research.
Batch extraction results can be used to determine whether or not measures to reduce nitrate leaching are effective or to determine trends. As batch extraction underestimates the nitrate concentration in water leaching from the root zone, it cannot be used directly for comparison with environmental quality objectives. Nevertheless, since the nitrate concentrations between both extraction methods were highly correlated, one might expect that nitrate concentrations observed from batch extractions might be used as a proxy for the nitrate concentration in leaching water as determined by centrifugation extraction. As such, batch extraction results can be used to clarify soil-type- and land-use-dependent spatial explicit patterns observed in national monitoring networks or to detect changes over time due to altered management and might even be useful to unravel the main drivers controlling the huge spatial and temporal variability in nitrate concentrations. Up to now, the estimate of the nitrate concentration in a centrifugation extract based on an individual batch result has been very uncertain (see Figure 7), but extending this approach for a wide range of nitrate concentrations and soil properties controlling the aggregate formation, as well as the AEC might help to build robust empirical models for regional applications. Nitrate concentrations in water leaching from the root zone of an agricultural field show inevitably a high variability, within a field, between fields on a farm, and between farms [71,72]. Consequently, a large number of samples is required to obtain a mean value with a small confidence interval for the area of interest. In such a case, the confidence interval of the estimated nitrate concentration in a centrifugation extract based on the results of a batch extraction will also be relatively small.
We did not find a significant contribution of SMR, soil moisture content, or CEC to the estimation of the centrifugation extract nitrate concentration, but SMR and CEC contributed to the estimation of the chloride concentration. For other soils or in the case of the use of other extraction solutions, it is advised to determine the soil-solution-specific relationship between centrifugation and batch extraction.

5. Conclusions and Outlook

This study showed that the nitrate concentration in the soil moisture differs within the soil matrix due to anion exclusion, thereby limiting the use of batch extractions for reliable estimates of the nitrate and chloride concentration actually leaching from agricultural soils. This is particularly relevant for loess soils that are sensitive to nitrate leaching. To monitor the effectiveness of measures improving groundwater and surface water quality and to guide policies and farmers’ practices, batch extracts can deliver “early warning” indicators. Batch extracts can also help to improve the spatial interpolation of the analysed data from centrifuged soil samples. Batch extraction of silt loams underestimates the nitrate and chloride concentration in the root zone leachate, as the concentration is higher in the mobile pore water than in the immobile pore water. By contrast, batch extraction may overestimate the concentration of nutrients that occur as precipitate in the soil matrix, such as sulphate, or that are in strong equilibrium with the cation exchange complex, such as ammonium. Nevertheless, batch extraction can be an inexpensive and reliable alternative to centrifugation to monitor relative changes over time and to estimate whether or not (policy) measures to reduce nitrate leaching in agriculture are effective. By establishing an empirical relationship between the concentrations in batch and centrifugation extracts, the concentration in the root zone leachate can be estimated via batch extracts.

Author Contributions

Conceptualisation, D.F.; methodology, D.F. and G.H.R.; software, D.F. and T.B.; validation, D.F., G.H.R. and T.B.; formal analysis, D.F. and T.B.; investigation, D.F.; resources, D.F.; data curation, D.F.; writing—original draft preparation, D.F.; writing—review and editing, G.H.R. and T.B.; visualisation, D.F.; supervision, D.F.; project administration, D.F.; funding acquisition, D.F. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the Ministerie van Landbouw, Natuur en Voedselkwaliteit (Ministry of Agriculture, Nature and Food Quality), as part of the National Minerals Policy Programme (Project Numbers M/350701/22 and M/350601/18).

Data Availability Statement

The data presented in this study are available upon request from the corresponding author. The data are not publicly available due to the protection of the individual privacy of the famers, but are available from the corresponding author upon reasonable request.

Acknowledgments

This research was carried out within the framework of the Minerals Policy Monitoring Programme. The authors thank the farmers for granting access to their fields, the RIVM field-work team for taking the samples, and the TNO laboratory team for carrying out the experiments. We would also like to thank Susanne Wuijts and Jappe Beekman for their critical reading of draft versions of the paper.

Conflicts of Interest

The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of the data; in the writing of the manuscript; nor in the decision to publish the results.

Appendix A

This appendix gives the change in the solute ratios in the centrifugation extracts with depth per sampling point for each of the 14 fields sampled in Experiment 2. It considers the nitrate-to-chloride, sulphate-to-chloride, and chloride-to-moisture-content ratios.
Water 15 02709 g0a1 550
Figure A1. Nitrate-to-chloride ratios with depth per sampling point for 14 fields. Fields are divided over four subplots (A–D) for visual clarity.
Figure A1. Nitrate-to-chloride ratios with depth per sampling point for 14 fields. Fields are divided over four subplots (A–D) for visual clarity.
Water 15 02709 g0a1
Water 15 02709 g0a2 550
Figure A2. Sulphate-to-chloride ratios with depth per sampling point for 14 fields. Fields are divided over four subplots (A–D) for visual clarity.
Figure A2. Sulphate-to-chloride ratios with depth per sampling point for 14 fields. Fields are divided over four subplots (A–D) for visual clarity.
Water 15 02709 g0a2
Water 15 02709 g0a3 550
Figure A3. Chloride-to-moisture-content ratios with depth per sampling point for 14 fields. Fields are divided over four subplots (A–D) for visual clarity.
Figure A3. Chloride-to-moisture-content ratios with depth per sampling point for 14 fields. Fields are divided over four subplots (A–D) for visual clarity.
Water 15 02709 g0a3

References

  1. Schulte-Uebbing, L.; De Vries, W. Reconciling food production and environmental boundaries for nitrogen in the European Union. Sci. Total Environ. 2021, 786, 147427. [Google Scholar] [CrossRef]
  2. Ahmed, M.; Rauf, M.; Mukhtar, Z.; Saeed, N.A. Excessive use of nitrogenous fertilizers: An unawareness causing serious threats to environment and human health. Environ. Sci. Pollut. Res. 2017, 24, 26983–26987. [Google Scholar] [CrossRef] [PubMed]
  3. Bryan, N.S.; Van Grinsven, H. The Role of Nitrate in Human Health. Adv. Agron. 2013, 119, 153–182. [Google Scholar] [CrossRef]
  4. United Nations. UN Paris Agreement. 2015. Available online: https://unfccc.int/files/essential_background/convention/application/pdf/english_paris_agreement.pdf (accessed on 22 June 2023).
  5. EC Farm to Fork Strategy. For a Fair, Healthy and Environmentally-Friendly Food System. European Commission, Brussels. 2020. Available online: https://food.ec.europa.eu/system/files/2020-05/f2f_action-plan_2020_strategy-info_en.pdf (accessed on 22 June 2023).
  6. Spijker, J.; Fraters, D.; Vrijhoef, A. A machine learning based modelling framework to predict nitrate leaching from agricultural soils across the Netherlands. Environ. Res. Commun. 2021, 3, 45002. [Google Scholar] [CrossRef]
  7. Di Bonito, M.; Breward, N.; Crout, N.; Smith, B.; Young, S.D.; Zhang, H. Extraction and Characterization of Pore Water in Contaminated Soils. In Environmental Geochemistry: Site Characterization, Data Analysis and Case Histories, 2nd ed.; De Vivo, B., Belkin, H.E., Lima, A., Eds.; Elsevier: Amsterdam, The Netherlands, 2018; Chapter 10; pp. 195–235. [Google Scholar]
  8. Fares, A.; Deb, S.K.; Fares, S. Review of vadose zone soil solution sampling techniques. Environ. Rev. 2009, 17, 215–234. [Google Scholar] [CrossRef]
  9. Weihermüller, L.; Siemens, J.; Deurer, M.; Knoblauch, S.; Rupp, H.; Göttlein, A.; Pütz, T. In situ soil water extraction: A review. J. Environ. Qual. 2007, 36, 1735–1748. [Google Scholar] [CrossRef]
  10. Ramos, C.; Kücke, M. A review of methods for nitrate leaching measurement. Acta Hortic. 2001, 563, 259–266. [Google Scholar] [CrossRef]
  11. Litaor, M.I. Review of soil solution samplers. Water Resour. Res. 1988, 24, 727–733. [Google Scholar] [CrossRef]
  12. Ji, W.; Huang, Y.; Li, B.; Li, Z. Identifying multivariate controls of water and nitrate in deep loess deposits under different land use types. J. Hydrol. 2022, 613, 128409. [Google Scholar] [CrossRef]
  13. Huang, P.; Zhang, J.; Zhu, A.; Li, X.; Ma, D.; Xin, X.; Zhang, C.; Wu, S.; Garland, G.; Pereira, E.I.P. Nitrate accumulation and leaching potential reduced by coupled water and nitrogen management in the Huang-Huai-Hai Plain. Sci. Total Environ. 2018, 610–611, 1020–1028. [Google Scholar] [CrossRef]
  14. Sprenger, M.; Erhardt, M.; Riedel, M.; Weiler, M. Historical tracking of nitrate in contrasting vineyards using water isotopes and nitrate depth profiles. Agric. Ecosyst. Environ. 2016, 222, 185–192. [Google Scholar] [CrossRef]
  15. Scanlon, B.R.; Gates, J.B.; Reedy, R.C.; Jackson, W.A.; Bordovsky, J.P. Effects of irrigated agroecosystems: 2. Quality of soil water and groundwater in the southern High Plains, Texas. Water Resour. Res. 2010, 46, W09538. [Google Scholar] [CrossRef] [Green Version]
  16. Parks, J.M.; Stewart, M.A.; McCartney, J.S. Validation of a centrifuge permeameter for investigation of transient infiltration and drainage flow processes in unsaturated. Geotech. Test. J. 2012, 35, 11. [Google Scholar] [CrossRef]
  17. Gillman, G.P. A Centrifuge Method for Obtaining Soil Solution; CSIRO report no. 16, 6; Commonwealth Scientific and Industrial Research Organization: Division of Soils: Canberra, Australia, 1976. [CrossRef]
  18. Moncur, M.C.; Blowes, D.W.; Ptacek, C.J. Pore-water extraction from the unsaturated and saturated zones. Can. J. Earth Sci. 2013, 50, 1051–1058. [Google Scholar] [CrossRef]
  19. Di Bonito, M. Trace Element in Soil Pore Water—A Comparison of Sampling Methods. Ph.D. Thesis, University of Nottingham, Nottingham, UK, May 2005. Available online: https://eprints.nottingham.ac.uk/10123/1/Marcello_Di_Bonito.pdf (accessed on 22 June 2023).
  20. Li, K.-Y.; Zhao, Y.-Y.; Yuan, X.-L.; Zhao, H.-B.; Wang, Z.-H.; Li, S.-X.; Malhi, S.S. Comparison of Factors Affecting Soil Nitrate Nitrogen and Ammonium Nitrogen Extraction. Soil Sci. Plant Anal. 2012, 43, 571–588. [Google Scholar] [CrossRef]
  21. Van Erp, P.J.; Houba, V.J.G.; Van Beusichem, M.L. Effect of drying temperature on amount of nutrient elements extracted with 0.01 M CaCl2 soil extraction procedure. Commun. Soil Sci. Plant Anal. 2001, 32, 33–48. [Google Scholar] [CrossRef]
  22. Orlowski, N.; Pratt, D.L.; McDonnell, J.J. Intercomparison of soil pore water extraction methods for stable isotope analysis and interpretation of hillslope runoff sources. Hydrol. Process. 2019, 33, 2939–2954. [Google Scholar] [CrossRef] [Green Version]
  23. Orlowski, N.; Pratt, D.L.; McDonnell, J.J. Intercomparison of soil pore water extraction methods for stable isotope analysis. Hydrol. Process. 2016, 30, 3434–3449. [Google Scholar] [CrossRef] [Green Version]
  24. Fraters, D.; Boom, G.J.F.L.; Boumans, L.J.M.; De Weerd, H.; Wolters, M. Extraction of soil solution by drainage centrifugation—Effects of centrifugal force and time of centrifugation on soil moisture recovery and solute concentration in soil moisture of loess subsoils. Environ. Monit. Assess. 2017, 189, 83. [Google Scholar] [CrossRef] [Green Version]
  25. Di Bonito, M.; Breward, N.; Crout, N.; Smith, B.; Young, S. Overview of selected soil pore water extraction methods for the determination of potentially toxic elements in contaminated soils: Operational and technical Aspects. In Environmental Geochemistry: Site Characterisation, Data Analysis and Case Histories, 1st ed.; De Vivo, B., Belkin, H.E., Lima, A., Eds.; Elsevier: London, UK, 2008; pp. 213–249. [Google Scholar] [CrossRef] [Green Version]
  26. Regelink, I.C.; Koopmans, G.F. Effects of Biostimulants and Fertilization on Nutrient Uptake by Grass and Composition of Soil Pore Water Versus 0.01 M CaCl2 Soil Extracts. Commun. Soil Sci. Plant Anal. 2021, 52, 2516–2532. [Google Scholar] [CrossRef]
  27. Moon, J.-W.; Paradis, C.J.; Joyner, D.C.; Von Netzer, F.; Majumder, E.L.; Dixon, E.R.; Podar, M.; Ge, X.; Walian, P.J.; Smith, H.J.; et al. Characterization of subsurface media from locations up- and down-gradient of a uranium-contaminated aquifer. Chemosphere 2020, 255, 126951. [Google Scholar] [CrossRef] [PubMed]
  28. Van den Brink, C.; Hoogendoorn, M.; Verloop, K.; De Vries, A.; Leendertse, P. Effectiveness of voluntary measures to reduce agricultural impact on groundwater as a source for drinking water: Lessons learned from cases in the Dutch provinces Overijssel and Noord-Brabant. Water 2021, 13, 3278. [Google Scholar] [CrossRef]
  29. Fraters, B.; Hooijboer, A.E.J.; Vrijhoef, A.; Plette, A.C.C.; Van Duijnhoven, N.; Rozemeijer, J.C.; Gosseling, M.; Daatselaar, C.H.G.; Roskam, J.L.; Begeman, H.A.L. Agricultural Practice and Water Quality in the Netherlands: Status (2016–2019) and Trend (1992–2019). Nitrate Report with the Results of the Monitoring of the Effects of the EU Nitrates Directive Action Programmes; RIVM report 2020–0184; National Institute for Public Health and the Environment: Bilthoven, The Netherlands, 2021. [CrossRef]
  30. Ma, B.; Huang, T.; Li, J.; Li, Z.; Long, Y.; Zhang, F.; Pang, Z. Tracing nitrate source and transformation in a semiarid loess aquifer with the thick unsaturated zone. Catena 2021, 198, 105045. [Google Scholar] [CrossRef]
  31. Huang, Y.; Evaristo, J.; Li, Z. Multiple tracers reveal different groundwater recharge mechanisms in deep loess deposits. Geoderma 2019, 353, 204–212. [Google Scholar] [CrossRef]
  32. Gates, J.B.; Scanlon, B.R.; Mu, X.M.; Zhang, L. Impacts of soil conservation on groundwater recharge in the semi-arid Loess Plateau, China. Hydrogeol. J. 2011, 19, 865–875. [Google Scholar] [CrossRef]
  33. Ma, B.L.; Ying, J.; Balchin, D. Impact of sample preservation methods on the extraction of inorganic nitrogen by potassium chloride. J. Plant Nutr. 2005, 28, 785–796. [Google Scholar] [CrossRef]
  34. Ros, G.H.; Hoffland, E.; Van Kessel, C.; Temminghoff, E.J.M. Extractable and dissolved soil organic nitrogen—A quantitative assessment. Soil Biol. Biochem. 2009, 41, 1029–1039. [Google Scholar] [CrossRef]
  35. Darrouzet-Nardi, A.; Weintraub, M.N. Evidence for spatially inaccessible labile N from a comparison of soil core extractions and soil pore water lysimetry. Soil Biol. Biochem. 2014, 73, 22–32. [Google Scholar] [CrossRef]
  36. Hamidian Shoormasti, N.; Tabatabaei-Nezhad, S.A. A novel mechanistic anion exclusion model to investigate partially water-saturated transport in soils and shales: A case study of nitrate solution flow. Eur. J. Soil Sci. 2022, 73, e13194. [Google Scholar] [CrossRef]
  37. Allred, B.J.; Brown, G.O.; Martinez, L.R. Laboratory investigation of boundary condition impacts on nitrate anion exclusion in an unsaturated soil. Soil Sci. Soc. Am. J. 2015, 79, 742–756. [Google Scholar] [CrossRef]
  38. González-Delgado, A.M.; Shukla, M.K. Coupled transport of nitrate and chloride in soil columns. Soil Sci. 2011, 176, 346–355. [Google Scholar] [CrossRef]
  39. Polubesova, T.; Borisover, M. Two components of chloride anion exclusion volume in montmorillonitic soils. Colloids Surf. A Physicochem. Eng. Asp. 2009, 347, 175–179. [Google Scholar] [CrossRef]
  40. Smethurst, P.J.; Herbert, A.M.; Ballard, L.M. A paste method for estimating concentrations of ammonium, nitrate, and phosphate in soil solution. Aust. J. Soil Res. 1997, 35, 209–225. [Google Scholar] [CrossRef]
  41. Lindau, C.W.; Spalding, R.F. Major procedural discrepancies in soil extracted nitrate levels and nitrogen isotopic values. Ground Water 1984, 22, 273–278. [Google Scholar] [CrossRef]
  42. Herbel, M.J.; Spalding, R.F. Vadose zone fertilizer-derived nitrate and δ15N extracts. Ground Water 1993, 31, 376–382. [Google Scholar] [CrossRef]
  43. Alves, A.C.; de Souza, E.R.; de Melo, H.F.; Pinto, J.G.O.; de Andrade Rego Junior, F.E.; de Souza Junior, V.S.; Marques, F.A.; do Santos, M.A.; Schaffer, B.; Raj Gheyi, H. Comparison of solution extraction methods for estimating electrical conductivity in soils with contrasting mineralogical assemblages and textures. Catena 2022, 218, 106581. [Google Scholar] [CrossRef]
  44. Jones, D.L.; Willett, V.B.; Stockdale, E.A.; Macdonald, A.J.; Murphy, D.V. Molecular weight of dissolved organic carbon, nitrogen, and phenolics in grassland soils. Soil Sci. Soc. Am. J. 2012, 76, 142–150. [Google Scholar] [CrossRef]
  45. Ali, T.; Bylund, D.; Essén, S.A.; Lundström, U.S. Liquid extraction of low molecular mass organic acids and hydroxamate siderophores from boreal forest soil. Soil Biol. Biochem. 2011, 43, 2417–2422. [Google Scholar] [CrossRef]
  46. Schuwirth, N.; Hofmann, T. Comparability of and alternatives to leaching tests for the assessment of the emission of inorganic soil contamination. J. Soils Sediments 2006, 6, 102–112. [Google Scholar] [CrossRef]
  47. Strobel, B.W. Influence of vegetation on low-molecular-weight carboxylic acids in soil solution—A review. Geoderma 2001, 99, 169–198. [Google Scholar] [CrossRef]
  48. Van Dijk, P.M.; Kwaad, F.J.P.M. Runoff generation and soil erosion in small agricultural catchments with loess derived soils. Hydrol. Process. 1996, 10, 1049–1059. [Google Scholar] [CrossRef]
  49. European Commission. EC Soil Atlas of Europe, European Soil Bureau Network; European Commission, Office for Official Publications of the European Communities: Luxembourg, 2005; p. L-2995. Available online: https://esdac.jrc.ec.europa.eu/content/soil-atlas-europe (accessed on 26 June 2023).
  50. Annual Average Precipitation, Average Values for 1991–2020. Available online: https://www.knmi.nl/klimaat-viewer/kaarten/neerslag-verdamping/gemiddelde-hoeveelheid-neerslag/jaar/Periode_1991-2020 (accessed on 22 June 2023).
  51. Houba, V.J.G.; Temminghoff, E.J.M.; Gaikhorst, G.A.; Van Vark, W. Soil analysis procedures using 0.01 M calcium chloride as extraction reagent. Commun. Soil Sci. Plant Anal. 2000, 31, 1299–1396. [Google Scholar] [CrossRef]
  52. R Core Team. R: A Language and Environment for Statistical Computing; R Foundation for Statistical Computing: Vienna, Austria, 2022; Available online: https://www.R-project.org (accessed on 22 June 2023).
  53. Wickham, H.; Averick, M.; Bryan, J.; Chang, W.; D’Agostino McGowan, L.; François, R.; Grolemund, G.; Hayes, A.; Henry, L.; Hester, J.; et al. Welcome to the tidyverse. J. Open Source Softw. 2019, 4, 1686. [Google Scholar] [CrossRef] [Green Version]
  54. Pinheiro, J.; Bates, D. R Core Team Nlme: Linear and Nonlinear Mixed Effects Models, R Package Version 3.1-157. 2022. Available online: https://CRAN.R-project.org/package=nlme (accessed on 22 June 2023).
  55. Extract Prediction Band from Lme fit. Answered 21 January 2013 at 9:47 (Roland), Edited 8 June 2018 at 15:18 (Hugo). Available online: https://stackoverflow.com/questions/14358811/extract-prediction-band-from-lme-fit (accessed on 8 February 2023).
  56. Lenth, R. Emmeans: Estimated Marginal Means, Aka Least-Squares Means, R Package Version 1.7.0. 2021. Available online: https://CRAN.R-project.org/package=emmeans (accessed on 22 June 2023).
  57. Djurhuus, J.; Jacobsen, O.H. Comparison of ceramic suction cups and KCl extraction for the determination of nitrate in soil. Eur. J. Soil Sci. 1995, 46, 387–395. [Google Scholar] [CrossRef]
  58. Lord, E.I.; Shepherd, M.A. Developments in the use of porous ceramic cups for measuring nitrate leaching. Eur. J. Soil Sci. 1993, 44, 435–449. [Google Scholar] [CrossRef]
  59. Ranger, J.; Marques, R.; Jussy, J.-H. Forest soil dynamics during stand development assessed by lysimeter and centrifuge solutions. For. Ecol. Manag. 2001, 144, 129–145. [Google Scholar] [CrossRef]
  60. Funk, R.; Maidl, F.-X.; Fischbeck, G. Nitratabbau in tieferen bodenschichten süddeutscher ackerstandorte auf löß|Denitrification in deeper soil layers of arable land in southern Germany. Z. Fur Pflanzenernahr. Und Bodenkd. 1996, 159, 207–213. [Google Scholar] [CrossRef]
  61. Wang, Y.; Ji, H.; Wang, R.; Hu, Y.; Guo, S. Synthetic Fertilizer Increases Denitrifier Abundance and Depletes Subsoil Total N in a Long-Term Fertilization Experiment. Front. Microbiol. 2020, 11, 2026. [Google Scholar] [CrossRef]
  62. Müller, C.; Clough, T.J. Advances in understanding nitrogen flows and transformations: Gaps and research pathways. J. Agric. Sci. 2014, 152, S34–S44. [Google Scholar] [CrossRef]
  63. Rohe, L.; Apelt, B.; Vogel, H.J.; Well, R.; Wu, G.-M.; Schlüter, S. Denitrification in soil as a function of oxygen availability at the microscale. Biogeosciences 2021, 18, 64, 1185–1201. [Google Scholar] [CrossRef]
  64. Gvirtzman, H.; Ronen, D.; Magaritz, M. Anion exclusion during transport through the unsaturated zone. J. Hydrol. 1986, 87, 267–283. [Google Scholar] [CrossRef]
  65. Gvirtzman, H.; Gorelick, S.M. Dispersion and advection in unsaturated porous media enhanced by anion exclusion. Nature 1991, 352, 793–795. [Google Scholar] [CrossRef]
  66. Heinrichs, H.; Böttcher, G.; Brumsack, H.-J.; Pohlmann, M. Squeezed soil-pore solutes—A comparison to lysimeter samples and percolation experiments. Water Air Soil Pollut. 1996, 89, 189–204. [Google Scholar] [CrossRef]
  67. Long, Y.; Huang, T.; Zhang, F.; Li, Z.; Ma, B.; Li, Y.; Pang, Z. Origin of sulphate in the unsaturated zone and groundwater of a loess aquifer. Hydrol. Process. 2021, 35, e14166. [Google Scholar] [CrossRef]
  68. Bibiso, M.; Tadesse, A.; Assefa, M. Evaluation of Three Universal Extractants for the Determination of P, NO3 and K in Some Soils of Ethiopia. Adv. Life Sci. Technol. 2012, 6, 16–24. [Google Scholar]
  69. Hamdi, W.; Gamaoun, F.; Pelster, D.E.; Seffen, M. Nitrate sorption in an agricultural soil profile. Appl. Environ. Soil Sci. 2013, 2013, 597824. [Google Scholar] [CrossRef] [Green Version]
  70. Webster, C.P.; Shepherd, M.A.; Goulding, K.W.T.; Lord, E. Comparisons of methods for measuring the leaching of mineral nitrogen from arable land. Eur. J. Soil Sci. 1993, 44, 49–62. [Google Scholar] [CrossRef]
  71. Boumans, L.J.M.; Fraters, D.; Van Drecht, G. Nitrate leaching in agriculture to upper groundwater in the sandy regions of the Netherlands during the 1992–1995 period. Environ. Monit. Assess. 2005, 102, 225–241. [Google Scholar] [CrossRef] [PubMed]
  72. Fraters, B.; Boumans, L.J.M.; Van Leeuwen, T.C.; De Hoop, W.D. Monitoring nitrogen leaching for the evaluation of the Dutch minerals policy for agriculture in clay regions. Sci. World J. 2001, 1, 758–766. [Google Scholar] [CrossRef] [Green Version]
Water 15 02709 g001 550
Figure 1. Study area, sample locations of Experiment 1 (dark blue circles, A–D), 5 boreholes per location (green dots in blue circle), and areas were 14 fields sampled for Experiment 2 (light blue). Brown colours on the left map are high elevations; green colours on the map are low elevations.
Figure 1. Study area, sample locations of Experiment 1 (dark blue circles, A–D), 5 boreholes per location (green dots in blue circle), and areas were 14 fields sampled for Experiment 2 (light blue). Brown colours on the left map are high elevations; green colours on the map are low elevations.
Water 15 02709 g001
Water 15 02709 g002 550
Figure 2. Setup of extraction of soil moisture by centrifugation and batch extraction in Experiment 1 (left) and Experiment 2 (right). Green oval shapes show the measured concentration in centrifugation (Cc) and batch extracts (Cb) and calculated concentrations in the batch extract of the centrifuged sample (C0) and the fresh soil sample (C0 fresh).
Figure 2. Setup of extraction of soil moisture by centrifugation and batch extraction in Experiment 1 (left) and Experiment 2 (right). Green oval shapes show the measured concentration in centrifugation (Cc) and batch extracts (Cb) and calculated concentrations in the batch extract of the centrifuged sample (C0) and the fresh soil sample (C0 fresh).
Water 15 02709 g002
Water 15 02709 g004 550
Figure 4. Ratio between nitrate and chloride (A) and sulphate and chloride (B) in a centrifugation extract versus a 1:1 Milli-Q water batch extract using soil measurements using samples from Locations A to D. Concentrations are expressed in mmol/L for the calculations. Confidence (Conf.Int) and prediction (Pred.Int) intervals are shown in dark and light pink, respectively.
Figure 4. Ratio between nitrate and chloride (A) and sulphate and chloride (B) in a centrifugation extract versus a 1:1 Milli-Q water batch extract using soil measurements using samples from Locations A to D. Concentrations are expressed in mmol/L for the calculations. Confidence (Conf.Int) and prediction (Pred.Int) intervals are shown in dark and light pink, respectively.
Water 15 02709 g004
Water 15 02709 g005 550
Figure 5. Effect of soil moisture recovery (SMR) on the difference in concentration between centrifugation and 1:1 Milli-Q water batch extracts for chloride (A) and sulphate (B) using samples from Locations A to D. Confidence (Conf.Int) and prediction (Pred.Int) intervals are shown in dark and light pink, respectively.
Figure 5. Effect of soil moisture recovery (SMR) on the difference in concentration between centrifugation and 1:1 Milli-Q water batch extracts for chloride (A) and sulphate (B) using samples from Locations A to D. Confidence (Conf.Int) and prediction (Pred.Int) intervals are shown in dark and light pink, respectively.
Water 15 02709 g005
Water 15 02709 g006 550
Figure 6. Effect of cation exchange capacity (CEC) on the difference in concentration between centrifugation and 1:1 Milli-Q water batch extracts for chloride using samples from Locations A to D. Confidence (Conf.Int) and prediction (Pred.Int) intervals are shown in dark and light pink, respectively.
Figure 6. Effect of cation exchange capacity (CEC) on the difference in concentration between centrifugation and 1:1 Milli-Q water batch extracts for chloride using samples from Locations A to D. Confidence (Conf.Int) and prediction (Pred.Int) intervals are shown in dark and light pink, respectively.
Water 15 02709 g006
Water 15 02709 g007 550
Figure 7. Relationship between nitrate concentration in soil moisture determined by a 1:2 (v/v) batch extraction with 0.01 M CaCl2 solution and by centrifugation. Confidence and prediction intervals are shown in dark and light pink, respectively. The intercept and regression coefficient and their significance are given in Table 2.
Figure 7. Relationship between nitrate concentration in soil moisture determined by a 1:2 (v/v) batch extraction with 0.01 M CaCl2 solution and by centrifugation. Confidence and prediction intervals are shown in dark and light pink, respectively. The intercept and regression coefficient and their significance are given in Table 2.
Water 15 02709 g007
Water 15 02709 g008 550
Figure 8. Relationships between the relative difference (%) in nitrate concentration in soil moisture determined by centrifugation and by a 1:2 (v/v) batch extraction with 0.01 M CaCl2 solution and nitrate concentration in batch extract (A), the sampling depth (B), the soil moisture content (C), and the soil moisture recovery (D). Confidence interval in dark pink and prediction interval in light pink.
Figure 8. Relationships between the relative difference (%) in nitrate concentration in soil moisture determined by centrifugation and by a 1:2 (v/v) batch extraction with 0.01 M CaCl2 solution and nitrate concentration in batch extract (A), the sampling depth (B), the soil moisture content (C), and the soil moisture recovery (D). Confidence interval in dark pink and prediction interval in light pink.
Water 15 02709 g008
Table
Table 3. Comparison of mean concentrations of nitrate and ammonium (mg/L) in soil moisture determined by centrifugation and 1:2 (v/v) 0.01 M CaCl2 solution extraction and percentage of samples with concentration below LoD 1.
Table 3. Comparison of mean concentrations of nitrate and ammonium (mg/L) in soil moisture determined by centrifugation and 1:2 (v/v) 0.01 M CaCl2 solution extraction and percentage of samples with concentration below LoD 1.
ParameterCentrifugationBatchDifferencet-Value
mg/L% < LoDmg/L% < LoD
Nitrate90.5 (120)066.8 (86.6)3024.0 (34.0)3.49 (4.13)
Ammonium0.130<2.7100--
Note: 1 the number in parentheses gives the result if data below LoD are neglected.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Fraters, D.; Ros, G.H.; Brussée, T. Measuring Nitrate Leaching in the Vadose Zone of Loess Soils—Comparison of Batch Extraction and Centrifugation. Water 2023, 15, 2709. https://doi.org/10.3390/w15152709

AMA Style

Fraters D, Ros GH, Brussée T. Measuring Nitrate Leaching in the Vadose Zone of Loess Soils—Comparison of Batch Extraction and Centrifugation. Water. 2023; 15(15):2709. https://doi.org/10.3390/w15152709

Chicago/Turabian Style

Fraters, Dico, Gerard H. Ros, and Timo Brussée. 2023. "Measuring Nitrate Leaching in the Vadose Zone of Loess Soils—Comparison of Batch Extraction and Centrifugation" Water 15, no. 15: 2709. https://doi.org/10.3390/w15152709

APA Style

Fraters, D., Ros, G. H., & Brussée, T. (2023). Measuring Nitrate Leaching in the Vadose Zone of Loess Soils—Comparison of Batch Extraction and Centrifugation. Water, 15(15), 2709. https://doi.org/10.3390/w15152709

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop