Application of diffusive gradient in thin films technique for mercury bioavailability prediction in soil

List of Figures Figure 2.1 The accumulation of Hg in diffusive gradients in thin films DGT resin and earthworm tissue as a function of deployment time for aged soils.. 23 Figure 2.2 The

Dissertation for Degree of Ph.D

Application of diffusive gradient in thin films technique for mercury bioavailability prediction in soil

Nguyen Huu Viet

School of Earth Sciences and Environmental Engineering Gwangju Institute of Science and Technology (GIST)

2021

Application of diffusive gradient in thin films

technique for mercury bioavailability prediction in soil

Advisor: Professor Seunghee Han

By Nguyen Huu Viet School of Earth Sciences and Environmental Engineering Gwangju Institute of Science and Technology (GIST)

A thesis submitted to the faculty of Gwangju Institute of Science and

Technology in partial fulfillment of the requirement for the degree of Doctor of Philosophy in the School of Earth Sciences and Environmental Engineering

Gwangju, Republic of Korea

Approved by

Professor Seunghee Han

Committee Chair

Application of diffusive gradient in thin films technique for mercury bioavailability prediction in soil

Nguyen Huu Viet Accepted in partial fulfillment of the requirement of the degree of Doctor of

Philosophy

May 31st, 2021 Dissertation advisor _

Prof Seunghee Han Committee member

Prof Kyoung-Woong Kim Committee member

Prof Chang-Keun Kang Committee member

Prof Yongseok Hong Committee member

Prof Nguyen Phuoc Dan

PhD./EN

20144049

Nguyen Huu Viet Application of diffusive gradient in thin films technique for mercury bioavailability prediction in soil School of Earth Sciences and Environmental Engineering 2021 P127

Advisor: Seunghee Han (한승희)

Hg exposure tests were conducted using freshly prepared artificial soils with different peat moss concentrations of 5, 10, 15, and 20% and aged prepared soils with varying pH values of 4.6, 5.6, and 6.2 It is interesting to note that the Hg uptake rates by DGT and earthworms were considerably higher for fresh soils than for aged soils, while pore water (and acid-extractable) Hg concentrations were rather similar between the two types of soils DGT-measured Hg flux used to assess bioavailability in earthworm showed a strong positive correlation with steady-state Hg concentration in earthworm ([earthworm Hg] = 354(DGT−Hg flux) − 34, R2 = 0.88) The overall results indicate that DGT-measured Hg flux is a better tool than the conventional methods for predicting Hg bioavailability for earthworms inhabiting diverse types of soil In the third chapter, the critical soil characteristics affecting Hg bioavailability to the earthworm Eisenia fetida were evaluated using DGT technique in the metal-contaminated soils collected from Gumu Creek, a

tributary of the Hyeongsan River The correlation analysis showed water holding capacity is the key variable of soil properties related to Hg accumulation in the soil, earthworm, and binding gel Indeed the water-holding capacity played a dual role in the Gumu Creek deposits: increasing the soil Hg concentration and decreasing Hg bioavailability and leachability DGT–Hg flux showed a positive correlated with the Hg concentration in earthworms (r = 0.93) The results of this chapter proved that the DGT method is promising for predicting soil Hg bioavailability to the earthworm Eisenia fetida, and the water-holding capacity simultaneously regulates Hg availability to the DGT and the earthworms In the last chapter, the aging effect of Hg, one of the most important factors controlling Hg bioavailability in soil, was studied using diverse types of field soils Surface soil samples were collected from forestry, agriculture, and riverbank sites (Youngsan and Hyeongsan river) The soil samples were spiked with inorganic Hg(II) and incubated for 1, 3, 5, 8, 15, 25, 40,

60 and 90 days We found that, during the aging process, the proportions of Hg tended to migrate from mobile fraction to the stable binding fractions Effective Hg concentration (CE) was estimated

by DGT and DIFS model, as it represents bioavailable fraction of Hg in soils according to the previous literature The effect of Hg aging on CE was evaluated by aging rate constant (k1) obtained

by fitting the CE values using pseudo-first order kinetic model Partial Least Square regression (PLSR) model was used to relate various soil properties to the variations of k1 in those soils PLS model constructed by cross-validation and variable selection routines predicted 31% of k1 when applied to entire soil samples, but 73 – 92% of k1 was predicted when applied to specific soil type The results showed that variation of k1 was mainly predicted by soil pH and organic matter content The overall results indicate that the combination of DGT technique and PLSR method is a useful tool for evaluating the aging effects on bioavailability of Hg in various types of soils using selected soil properties

Table of Contents

Abstract i

List of Figures vii

List of Tables x

1.1 Introduction 1

1.1.1 Mercury cycling in the environment 1

1.1.2 Hg in soil ecosystem 2

1.1.3 Bioavailability and toxicity of Hg in soil 3

1.1.4 DGT technique for Hg measurement 4

1.2 DGT technique to evaluate lability of Hg in soil 6

1.2.1 DGT application for labile Hg measurement in soil 6

1.2.2 DGT technique as a bio-mimics surrogate for mercury bioavailability in soil 7

1.2.3 Modelling approaches for predicting the bioavailability of Hg in soil using DGT 8

Chapter 2: DGT efficacy tests using earthworm Eisenia fetida grown in artificial soils 12

Abstract 12

2.1 Introduction 13

2.2 Experimental methods 15

2.2.1 DGT probe preparation 15

2.2.2 Preparation and characterization of soils 16

2.2.3 Deployment of DGT probes and earthworms 19

2.2.4 Hg measurement in DGT probes, earthworms, and soils 19

2.2.5 Calculation of CDGT and OCM simulation 21

2.3 Results and discussion 22

2.3.1 Effect of Hg concentration on earthworm and DGT accumulation of Hg 22

2.3.2 Effect of peat moss content on earthworm and DGT accumulation of Hg 26

2.3.3 Effect of soil pH on earthworm and DGT accumulation of Hg 30

2.3.4 Effect of soil aging on earthworm and DGT accumulation of Hg 31

2.3.5 Prediction of Hg bioavailability using DGT and conventional methods 34

2.4 Conclusions 36

Chapter 3: Applying DGT technique for evaluating Hg bioavailability by earthworm Eisenia fetida in natural soils 38

Abstract 38

3.1 Introduction 39

2.2 Materials and Methods 42

2.2.1 DGT unit and earthworm preparation 42

2.2.2 Soil sampling and pre-treatment 43

2.2.3 Deployment of earthworms and DGT units 45

2.2.4 Measurement of soil characteristics 46

2.2.5 Measurement of Hg in resins, earthworms, soils, and headspace air 48

2.2.6 Accumulation factors and statistical analysis 51

2.3 Results and Discussion 52

2.3.1 Soil properties 52

2.3.2 Hg pollution in soils 55

2.3.3 Hg accumulation in earthworms 58

2.3.4 Hg accumulation in DGT 62

2.3.5 Prediction of BAF and DAF using PLSR 64

2.4 Conclusions 67

Chapter 4: Applying DGT techniques for evaluating aging effects of Hg in natural soils 69

Abstract 69

4.1 Introduction 70

4.2 Materials and methods 71

4.2.1 Soil sampling and preparation 72

4.2.2 Incubation method 73

4.2.3 DGT fabrication 74

4.2.4 Measurement of Hg in binding layer and soils 75

4.2.5 Soil properties measurements 75

4.2.6 Effective concentration calculation 78

4.2.7 Kinetic equations for aging and leaching process 79

4.2.8 Statistical analysis 80

4.3 Results and discussion 80

4.3.1 Soil characterization 80

4.3.2 Correlation analysis among soil properties 83

4.3.3 Aging effect of Hg on DGT availability and the role of soil properties 85

4.3.4 Prediction of aging rate k1 and PAF using PLS method 88

4.4 Conclusions 92

Chapter 5: Conclusion 94

References 95

List of Figures

Figure 2.1 The accumulation of Hg in diffusive gradients in thin films (DGT) resin and earthworm tissue as a function of deployment time for aged soils The experimental Hg data were fitted by one compartment model (OCM) Soil composition was 69% sand, 20% kaolinite, 10% peat moss, and 1% CaCO3, and the measured Hg concentration was A 5.2 nmol g-1, B 56 nmol g-1, C 105 nmol g-1, and D 262 nmol g-1 23

Figure 2.2 The accumulation of Hg in diffusive gradients in thin films (DGT) resin and earthworm tissue as a function of deployment time for fresh soils The experimental Hg data were fitted by one compartment model (OCM) Soil composition was 59-74% sand, 20%

kaolinite, and 1% CaCO3, and the peat moss content was A 5%, B 10%, C 15%, and D 20% 29

Figure 2.3 The accumulation of Hg in diffusive gradients in thin films (DGT) resin and earthworm tissue as a function of deployment time for fresh soils The experimental Hg data were fitted by one compartment model (OCM) Soil composition was 69% sand, 20% kaolinite, and 10% CaCO3, and the pH of the soil was A 4.6, B 5.6, and C 6.2 30

Figure 2.4 A) Hg in earthworm tissue at steady-state and B) Hg uptake rate by

earthworm versus porewater Hg concentration The Hg uptake rate was estimated by k1×Cs, where Cs is the soil Hg concentration and k1 is the Hg uptake rate constant for earthworm

estimated by the one compartment model (OCM), as presented in Table 2 32

Figure 2.5 A) Hg in earthworm tissue at steady-state and B) Hg uptake rate by

earthworm versus acid-extractable Hg concentration The Hg uptake rate was estimated by

k1×Cs, where Cs is the soil Hg concentration and k1 is the Hg uptake rate constant for earthworm estimated by the one compartment model (OCM), as presented in Table 2 33

Figure 2.6 Correlation of A) Hg in earthworm tissue at steady-state and B) Hg uptake

rate by earthworm with diffusive gradients in thin films (DGT)-measured Hg flux at 24 hours of deployment time The Hg uptake rate was estimated by k1×Cs, where Cs is the soil Hg

concentration and k1 is the Hg uptake rate constant for earthworm estimated by the one

compartment model (OCM), as presented in Table 2 35

Figure 3.1 The location of the sampling sites in Gumu Creek in Pohang, South Korea 44 Figure 3.2 (a) A setting to measure the accumulation of Hg(0) in a headspace and DGT resin above the air film and (b) Hg(0) concentration accumulated in a headspace after 4 and 6 weeks of incubation at 20℃ Hg concentrations in the DGT resins were under the detection limit

Soil Hg concentrations are found from Table 3 50

Figure 3.3 The correlation between Hg concentration in earthworm (HgEW) and measured Hg flux The Hg concentrations in earthworms were determined after 6 weeks of

DGT-earthworm incubation in soils and the DGT-Hg fluxes were measured after 24 hours of DGT deployment after the earthworm collection 64

Figure 3.4 The partial least squares regression projection of the independent variables (soil properties) and dependent variables (BAF and DAF) The variable importance in the

projection (VIP) index of input parameters are found in Table 4 WHC: water holding capacity, CEC: cation exchange capacity, LOI: loss on ignition, and ORP: oxidation reduction potential 65

Figure 4.1 Location of sampling sites 72 Figure 4 2 Grain size classification plot of 7 forest, 6 agricultural, 5 Youngsan river and

3 Hyeongsan river soils 82

Figure 4.3 Changes in effective concentration (CE) of forest, agricultural, Youngsan river, and Hyeongsan river soils 87

Figure 4.4 Predictor (red dots) and response (blue dots) loadings on factor 1 and 2 (T1

and T2) in initial partial least squares regression (PLSR) model for k1, 𝐶𝐸0 and PAF in a) entire soil samples, b) forest soils, c) agricultural soils, d) Youngsan river soils and e) Hyeongsan river soils 89

List of Tables

Table 2.1 Composition and chemical characteristics of artificial soils for diffusive

gradients in thin films (DGT) and earthworm deployment: pH, total organic carbon (TOC), aging period, water content, and maximum water holding capacity (MWHC) 18

Table 2.2 Output parameters of the one compartment model [i.e., diffusive gradients in thin films (DGT) and earthworm uptake rate constant, elimination rate constant, DGT-soil

accumulation factor (DSAF), and biota-soil accumulation factor (BSAF)], and experimentally determined DSAF and BSAF 25

Table 3.1 Conventional properties of the soil samples collected from the Gumu Creek, a tributary of Hyeongsan River CEC: cation exchange capacity, LOI: loss on ignition, WHC: water holding capacity, and ORP: oxidation reduction potential The values are the mean and standard deviation of triplicate measurements 54

Table 3.2 Pearson’s correlation matrix of the Hg concentrations in soil and earthworm, DGT-Hg flux, bioaccumulation factor (BAF), DGT accumulation factor (DAF), and general properties of the riverbank soils collected from the Gumu Creek WHC: water holding capacity, LOI: loss on ignition, ORP: oxidation reduction potential, and CEC: cation exchange capacity P

< 0.05 is shown in bold 56

Table 3.3 Summary of Hg concentrations in soils, geo-accumulation index (Igeo), Hg concentrations in earthworms, bioaccumulation factor (BAF), Hg concentrations in DGT resin, DGT-measured Hg flux, and DGT accumulation factor (DAF) of the soil samples collected from the Gumu Creek The values are the mean and standard deviation of triplicate measurements Soil quality was assessed based on the Soil Pollution Concern Guidelines set by the Ministry of Environment in South Korea 59

Table 3.4 The partial least squares regression equations constructed to simulate the log bioaccumulation factor (BAF) and log DGT accumulation factor (DAF) using soil properties Variable importance in the projection (VIP), root mean square error (RMSE), and normalized root mean square error (NRMSE) are shown in the table WHC: water holding capacity, ORP: oxidation reduction potential, CEC: cation exchange capacity 65

Table 4.1 Geographical coordinates of the soil sampling sites 73 Table 4.2 Summary statistics of soil properties of forest, agricultural, Youngsan river and Hyeongsan river soils Cation Exchange Capacity (CEC) (meq/100g), Water holding

capacity (WHC) (%), pHCaCl2, pHH2𝑂, Sand (%), Silt (%), Clay (%), ORP( mV), LOI (%), TC (%), TOC (%), TN (%), TS (%), dithionite extracted Iron and Aluminum (Fed and Ald)(%), pyrophosphate extracted Iron and Aluminum (Fep and Alp)(%), Hydroxylamine extracted Iron and Aluminum (Feh and Alh)(%) Error! Bookmark not defined

Table 4.3 Pearson’s correlation matrix of soil properties Error! Bookmark not defined.Table 4.4 Data for nonlinear modelling of adsorption kinetics using 1st and 2nd pseudo-order models 85

Table 4.5 The partial least squares regression equations constructed to simulate k1, PAF and CE0 using soil properties selected by Variable Importance in the Projection (VIP) method with VIP>1 Root mean square error (RMSE) was shown in the table Multiple linear regression

by PLSR method was applied for entire soil, forest, agricultural, Youngsan river and Hyeongsan river soils analysis 90

Chapter 1: Introduction of the diffusive gradient in thin films (DGT) technique for passive sampling of Hg in soil

1.1 Introduction

1.1.1 Mercury cycling in the environment

Mercury (Hg) is one of the top ten concern elements listed by WHO 2019 Hg presents ubiquitously as a contaminant at low concentration in all the compartments of the environment (Driscoll et al 2013, Xu et al 2015, Yang et al 2020, Clarkson et al 2003) Hg releases into the atmosphere, aquatic, and terrestrial ecosystems, derived from various natural (i.e., volcano eruption, forest fires, cinnabar) and anthropogenic activities (i.e., fuel combustion, industrial processes, and the chlor-alkali industry) (Obrist et al 2018, Amos et al 2015, Streets et al 2018, Gworek et al 2020) The expansion of worldwide industrialization after World War II drove a significant increase of the emission of Hg to the environment (Streets et al 2017) In the recent update by UNEP (2003), approximately 30% of direct Hg emission to the atmosphere is from anthropogenic activities, while 10% the emission is related to natural sources The reemission of legacy Hg accounts for 60% of total emission, and it comes from previously anthropogenic sources built up for a long-term period on the surface of soil and ocean that return to the atmosphere Once entering to the environment, Hg can be transported in the atmosphere over very long distances due

to the long resident time of Hg, which is fundamental to Hg global cycling (Schroeder et al 1991, O'Connor et al 2019, Kim &Zoh 2012) Mercury cycles in the terrestrial, aquatic, and atmospheric ecosystem, until it moved out of the system and buried in soil and sediment through entrapment in stable mineral compounds (Obrist et al 2018, Bishop et al 2020, Selin 2009) The estimation of global anthropogenic Hg emission to the atmosphere was approximately 2220 tons per year,

including emission (UNEP 2019) UNEP (2008) estimated the sum of natural emission and emission to be in the range of 1,800 – 4,800 tons yr-1 The amount of Hg mass accumulated in soils

re-is very large, assumed to be in the range of 250 – 1000 Gg (Obrre-ist et al 2018) A significant proportion of Hg in soils is attributed to anthropogenic influences, with an estimated 86 Gg of anthropogenic Hg is now accumulated in surface soils (Futsaeter &Wilson 2013)

Mercury could be found in the environment as Hg0 (elemental mercury), Hg+1 (mercurous) and Hg+2 (mercuric)( Bigham et al 1964, Park &Zheng 2012, Bernhoft 2012) In soils, Hg0 and

Hg+2 compounds are stable in typical conditions, of these, Hg+2 is the predominant oxidation state

of mercury compounds, while Hg (0) is very susceptible to oxidation Hg(0) volatilizes easily to the atmosphere (Teng et al 2020, Gai et al 2016, Jagtap &Maher 2015) In the atmosphere, the elemental Hg, (Hg(0) account for 98% of the total species (Cohen et al 2004, Bloom &Fitzgerald

1988, Morel et al 1998), while in the aquatic ecosystem, the Hg(0) concentration is high variable depend on the depth, location, season, bacteria, and plankton (Morel et al 1998, Jiang et al 2018)

Hg (0) was discovered in higher concentrations closer to the air-water interface whereas total Hg and methylmercury (MeHg) exist in greater concentrations near the sediment-water interface (Morel et al 1998,Gill et al 1999) In sediments, Hg is mainly bound to sulphur as well as organic matter and inorganic particles (Morel et al 1998, Hesterberg et al 2001, Zhong &Wang 2009)

industrial areas in Spain reported concentrations of Hg roughly 152 ± 47 μg kg−1 Fernandez et al 2014) In mining area in Almadén, Spain, high concentrations was detected around

(Gonzalez-9000 mg kg−1 (Higueras et al 2003) In soil in state of Virginia, America, Hg was found at 0.1–94

μg kg−1 in an industrial area (Bigham et al 2015) The high mercury concentrations are related to the proximity to natural or anthropogenic Hg sources

Soil is the largest terrestrial Hg pool and play a critical role in the global Hg cycling as a sink for atmospheric Hg deposition and a re-emission source (Jiskra et al 2015, Wang et al 2019) Several pathways lead to atmospheric Hg deposition to terrestrial surfaces Gaseous Hg(0) is oxidized to Hg(II) in the atmosphere and deposited to terrestrial surfaces directly through precipitation (Jiskra et al 2015, Yin et al 2014b) Once taken up by plant, Hg(II) can be photo-reduced to Hg(0) and re-emit to atmosphere (Jiskra et al 2015, Selin 2009) Under typical soil conditions, Hg(II) occur abundantly in the forms of inorganic mercuric salts (i.e., HgCl2, HgO, HgS) and minerals (primarily cinnabar, metacinnabar) (O'Connor et al 2019, Yin et al 2016) Hg(II) can transform under the influence of soil redox potential and exist dominantly in Hg-S compounds in oxidizing conditions Under the reducing condition, biological and chemical processes transform Hg+2 to organic mercury species mainly methylmercury (CH3Hg) which is the most toxic form of Hg In organic soils, the dominant Hg form is Hg(II), mainly bound to reduced sulfur groups of natural organic matter (NOM) In mineral soils, Hg(II) is primarily adsorbed to mineral surfaces such as Fe, Al, Mn oxides or clay minerals (references are needed here)

1.1.3 Bioavailability and toxicity of Hg in soil

Bioavailability of Hg in soil is related to labile Hg species, as it controls the uptake fraction

of Hg by living biota as well as Hg leaching through the soil (Wang et al 2014a) The concentration

of labile Hg means available Hg in the soil solution and resupplied Hg from the soil particles (Reis

et al 2015, Shetaya et al 2017) Bioavailability of Hg defines the correlation between the total Hg concentration in the soil and the level of Hg uptake into the biota (Fang et al 2011, Araujo et al 2019) Plant or terrestrial animals such as earthworm have been frequently used to evaluate the bioavailable pool of Hg in soil Wang et al (2014) revealed that the fraction of Hg in soil that can

be uptaken by plants directly reflects bioavailable Hg Strong and significant correlations were found between the Hg amounts in vegetables and the total Hg concentrations in farmland soils near

to fluorescent lamp factories in China (Shao et al 2012) Meanwhile, earthworm is the organism that has been widely used for investigating the bioavailability of Hg in terrestrial ecosystem, due

to its important as a major soil ecosystem component and a food source for variety of organisms Multiple field studies have been conducted to examine the bioaccumulation of Hg by individual earthworm species

1.1.4 DGT technique for Hg measurement

The diffusive gradients in thin films (DGT) is an in-situ sampling technique that was first applied by Davison &Zhang (1994) for the measurement of heavy metal in freshwater This technique was later applied to sediment and soil Piston and plate are the two common DGT device types that have been reported Piston type DGT was used for solution and soil application, while the plate-type DGT was used for sediment application (Ding et al 2016) The DGT device typically comprises a binding hydrogel layer embedded with specific adsorbent, a diffusion layer composed of a hydrogel, and a filter membrane The binding gel, diffusive gel, and filter membrane stack were laid together and sealed tightly by a plastic base and cap The difference in concentration between binding layer and DGT surface generates a concentration gradient, which

is a driving force that allow solutes to pass through the diffusive layer and accumulate in binding layer Initially, the application of the technique to Hg was failed by the fact that Hg accumulates

in amide groups of polyacrylamide gel working as the diffusive layer in original method The DGT method was modified by replacing the polyacrylamide gel by agarose (Divis et al 2005), which allow Hg to pass freely through with a diffusion coefficient close to water (Docekalova &Divis 2005) Variety of binding agents have been applied for Hg depending on the method and target species, such as Spheron-Thiol, 3-mercaptopropyl-functionalized silica gel (3-MFSG), Duolite GT73, Iontosorb AV-MP, and Ambersep GT74

According to the principles of the technique, analyte concentration in the solution can be determined by the equation below:

𝐶 = 𝑀 × ∆𝑔

𝐷 × 𝑡 × 𝐴 (1) where Cb is the free or labile concentration of metals in the deployment media, M is the recovered mass of the analyte, g is the diffusive gel thickness, D is the diffusion coefficient of the analyte in the gel, t is the deployment time, and A is the exposure window area Temperature dependent D can be calculated using Equation (2)

𝑙𝑜𝑔𝐷 = 1.37 × (𝑇 − 25) + 8.36 × 10 × (𝑇 − 25)

273 + 𝑇

298 (2) where D25

g is the diffusion coefficient of metal in a diffusive gel at 25 °C

Considering the thickness of the DBL (δ) and filter layer (Δf) (Garmo et al 2006; Kreuzeder et al 2015), Equation (1) can also be expressed as:

In equation (1), the accumulated mass M can be measured directly in the binding-gel layer

by combustion method (using AMA-254) or using a beam technique such as laser ablation ICPMS

or, in the case of radionuclides, by direct counting It can be also obtained via elution of the ion from the binding gel into solution for routine measurements to minimize the cost The accumulated mass, M, can be calculated using the following Eq 4

𝑀 =𝐶 (𝑉 + 𝑉 )

𝑓 (4) where Ce is the concentration of the ion in the elution solution, Ve is the volume of the elution solution, Vg is the volume of the gel and ƒe is the elution efficiency

1.2 DGT technique to evaluate lability of Hg in soil

1.2.1 DGT application for labile Hg measurement in soil

Mercury speciation and its interaction with environmental components to form soluble, colloidal, and Hg insoluble complexes affects to the mobility and bioavailability to organism The DGT method is advantageous for in-situ measurement of Hg speciation, because DGT preserves stability and distribution of the Hg species, while it also preconcentrates target analyte during the deployment Cattani et al (2008) evaluated the rhizosphere Hg bioavailability from contaminated soil by combine DGT and HPLC-ICP-MS methods The results indicated that both of the methods are appropriate for the estimation of Hg root availability from contaminated soils Cattani et al (2009) investigated the influence of humic acid (HA) on Hg mobility and bioavailability in contaminated soils by DGT method In soil, Hg can be mobilized or stabilized in binding sites of humic substances By using both ultrafiltration and DGT methods for assessing the lability of Hg-HA complexes, the author demonstrated that DGT can be used to estimate the labile fractions of Hg-HA complexes DGT technique is capable of measuring free Hg ions in soil solution, Hg from the dissociation of the complexes with organic and inorganic ligands and Hg

water-complexes released from the solid phase The DGT results showed a significant enhancement of

Hg availability in the presence of HA, which extracted Hg from soil particle to soil solution Turull

et al (2019b) successfully applied DGT to quantify the inorganic and organic Hg complexes which

is in the forms with humic and fulvic acids The author has used two diffusive layers with different pore size to distinguish the inorganic and organic bioavailable Hg species in soil The large-pore size (>5 nm) diffusive layer which is called open diffusive layer (ODL) could allow both of inorganic and organic labile Hg species to pass through, while small-pore size (<1nm) diffusive layer defined as restricted diffusive layer (RDL) only allow free ions and small inorganic Hg complexes to diffuse freely The results revealed that combining ODL and RDL improved the accuracy of DGT measurement in assessing bioavailability and toxicity of Hg in organic and inorganic forms separately In further study, Turull et al (2019a) applied DGT method with ODL and RDL for inorganic and organic Hg labile in amendment of agricultural soils by adding biochar and compost The amendment increased or reduced the bioavailability, mobility, and toxicity of

Hg The application of DGT-RDL pointed out an increase of inorganic Hg fractions from 16 to 20% with the increase of biochar from 3 to 6%, and a significant decrease from xx to 35% in the case of 30% compost amendment

1.2.2 DGT technique as a bio-mimics surrogate for mercury bioavailability in soil

The DGT technique has provided a reliable surrogate measurement to be a good predictor

of Hg uptake by plants DGT is likely to mimics the uptake processes of Hg by plant root The method is also advantageous for in situ measurement of Hg speciation because it preserves the stability and distribution of the Hg species, while also able to concentrate trace element during deployment Cattani et al (2008) applied DGT and HPLC–ICP-MS for evaluate the accumulation

of Hg in plant root The results confirmed both methods were suitable tool for estimating the Hg root availability and risk from contaminated soils Liu et al (2012) tested the ability of DGT method for predicting MeHg uptake by rice (Oryza sativa L.) A significant correlation between MeHg flux in soil measured by DGT and MeHg flux in rice roots was found (R = 0.853, p < 0.01) The author indicated the potentially prediction of the bioavailability of MeHg in rice paddy soil

by DGT method and quantitatively provide the rate of uptake of bioavailable MeHg Nguyen et al (2019) measured the bioavailability of Hg in soil for the earthworm Eisenia fetida by using DGT embedded with 3-mercaptopropyl-functionalized silica as the binding phase The bioavailable Hg

in soil varied in the effects of pH (4.6 – 6.2) and organic matter content (5 – 15%) during 10 days

of deployments The calculated uptake efficiency for Hg was 91 ± 3.4%, the DGT method showed effectively predict Hg concentration of earthworm tissue compared to conventional methods of pore water and 0.2M HNO3 extractable Hg

1.2.3 Modelling approaches for predicting the bioavailability of Hg in soil using DGT

Beside the application of DGT for determining labile Hg concentration, DGT-measured

Hg concentration was recently developed to predict bioavailability, based on the correlation between Hg in DGT and Hg in bio-indicator organism (i.e., plant, earthworm) Some studies have been published to evaluate the possibility to use DGT to predict bioavailability and toxicity in soil system (Turull et al 2019b, Nguyen et al 2019, Liu et al 2012a, Ridošková et al 2016) The method aims to determine the labile fraction of Hg in soil matrices and understand how the bioavailability changes with varied environmental parameters (e.g., pH, redox potential, organic matter content) DGT accumulation of Hg simulated Hg uptake by earthworm through dermal route, which account for 96% of total accumulated Hg (Le Roux et al 2016), or by plants through

the diffusion of labile Hg from the rhizosphere soil (Liu et al 2012a, Turull et al 2019b) The kinetic and equilibrium constants of the bioaccumulation of Hg in soils has been obtained using diverse simulation models, such as the one compartment model (Nguyen et al 2019), first-order dissociation kinetic (Liu et al 2012a), and DIFS model (Turull et al 2019a, Turull et al 2019b, Ridošková et al 2016,)

Liu et al (2012) have investigated the relevance of DGT to investigate the MeHg dynamic and spatial MeHg trend in pore water and tested the ability of DGT to predict MeHg uptake by rice The experiment results showed a significant correlation between DGT-measured MeHg flux and MeHg uptake in the root (R = 0.853, p<0.01) The author demonstrated that DGT could quantitatively provide the uptake rate of bioavailable MeHg and has a potential in predicting the bioavailability of MeHg in contaminated paddy soil Furthermore, the quantitative MeHg content

in grain could be estimated as a function of the DGT-measured flux of MeHg in soils by the equation

[𝑀𝑒𝐻𝑔] = [(36.14𝐹 − 1.69) × 10 ] × 𝑅𝑊𝑅 × 𝑆𝑅𝐴 × 𝑡 × 𝑘

where [MeHg]rice represents the total MeHg concentration in the rice of a single plant (ng

g–1), RWR is the ratio of root to whole plant biomass (g root/g plant), SRA is the specific root area (cm2 g–1), t is a set growing period (h), and k is the % of MeHg in the plant that is stored in rice

FDGT is the flux of MeHg measured by the DGT technique (ng MeHg cm–2 h–1)

Nguyen et al (2019) measured the bioavailability of Hg in laboratory-prepared artificial soils for the earthworm Eisenia fetida by using DGT which used 3-mercaptopropyl-functionalized silica as the binding phase The soils used in Hg exposure experiments were prepared with different

pH (4.6, 5.6, and 6.2) and varying peat moss concentrations of 5, 10, 15, and 20% DGT was deployed in the soil with approximately 35–40 earthworms and removed with a time period of

half-a-day to 10 days The calculated uptake efficiency for Hg for this binding gel was 91 ± 3.4%; the DGT was able to predict the Hg concentration of earthworm tissue effectively, unlike comparative measurements of pore water and acid-extractable Hg The research showed that 24h-integrated DGT-Hg flux is a better tool for predicting the Hg bioavailable in earthworm (R2 = 0.88) Hg uptake in earthworm is in linear correlation with DGT-Hg flux by equation:

[𝐻𝑔] = 354 × [𝐷𝐺𝑇 − 𝐻𝑔 𝑓𝑙𝑢𝑥] − 34 Turull et al (2019a) has performed the results of predicting bioavailability of soil Hg uptake by lettuce by using the DGT technique to detect bioavailable Hg in uncontaminated agricultural soil the prediction was conducted by evaluating the Hg uptake in lettuce and the effective concentration CE calculated by DIFS program from DGT-Hg results that theoretically same concentration found in rhizosphere of plants The author successfully predicts bioavailable inorganic Hg species and organic species on lettuce by DGT by applying open and restricted diffusive layers Similar results were reported by Turull et al (2019b) and Ridošková et al (2012) for applying DIFS calculated effective concentration (CE) to predict Hg bioavailability to plants in amendment agricultural soil and urban soil

Nguyen et al (2019) has conducted a research to assess the critical soil characteristics affecting on Hg bioavailability to the earthworm Eisenia fetida using DGT measurement in natural heavy Hg contaminated soils In this study, Hg concentration in earthworm tissue was predicted

by DGT-Hg flux with a positively high correlation (r = 0.93) Further, Hg bioaccumulation factor (BAF) and DGT- Hg accumulation factor (DAF) was successfully simulated by Partial Least Square Regression method The same study showed that log BAF ( r2 = 0.85) and log DAF (r2 = 0.97) can be simulated using diverse soil properties (i.e., water holding capacity, pH, loss of ignition, redox potential, cation exchange capacity, clay, silt, sand, total soil carbon, total nitrogen

and total sulfur) Overall, DGT is a reliable technique to measure the bioavailability of Hg in soil The DGT-Hg flux is a useful tool to predict Hg bioaccumulation in plant and terrestrial organism, while effective concentration (CE) is powerful in predicting Hg uptake by plants

Chapter 2: DGT efficacy tests using earthworm Eisenia fetida grown in artificial soils

Abstract

In general, the diffusive gradients in thin films (DGT) technique is an effective tool for evaluating metal bioavailability, however, its applicability is subject to the type of metal and organism involved In this study, the accumulated masses of Hg in DGT probes and in the earthworm species Eisenia fetida were monitored for 10 days, to test if DGT technique can be used

as a predicting method for the bioavailability of soil Hg to earthworms In the Hg exposure tests using soils prepared with different peat moss concentrations of 5, 10, 15, and 20% and varying pH values of 4.6, 5.6, and 6.2, the experimentally determined DGT-soil accumulation factor (DSAF) and biota-soil accumulation factor (BSAF) both increased as the peat moss content decreased, and the pH increased According to a one-compartment model, this was a result of the increased Hg uptake rate constant (k1) and the relatively stable Hg elimination constant (k2) under lower peat moss and higher pH conditions It is interesting to note that the Hg uptake rates by DGT and earthworms were considerably higher for fresh soils than for aged soils, while pore water (and acid-extractable) Hg concentrations were rather similar between the two types of soils Across diverse soil properties, steady-state Hg in earthworm tissue showed a strong positive correlation with DGT-measured Hg flux ([earthworm Hg] = 354(DGT-Hg flux) - 34, r2 = 0.88), while meager correlations were found between Hg concentration in earthworms and that in pore water (and acid-extractable) The overall results indicate that DGT-measured Hg flux is a better tool than conventional methods for predicting Hg bioavailability for earthworms inhabiting diverse types of soil

2.1 Introduction

Research efforts have been made to develop experimental tools for assessing the bioavailability of metals in soil and sediment (Desaules 2012) The conventional methods of predicting bioavailability include the sequential extraction, which measures chemically reactive metals with a series of solvents, and the equilibrium partitioning approach, which assesses the metal concentration based on the comparison between the dilute acid-extractable metal concentration and the concentration of acid-volatile sulfide (AVS) measured in the sediment In general, these techniques provide indications of metal bioavailability by measuring weakly bound metals on soil particles (Roig et al 2013; Ren et al 2015) However, the reported disadvantages

of both approaches include relatively weak inter-laboratory reproducibility and the fact that test results vary based on soil properties (De Jonge et al 2010; Yin et al 2014a)

The diffusive gradients in thin films (DGT) technique was designed to measure averaged concentrations of dissolved metals in soils (Zhang et al 2001; Clarisse et al 2009; Hong

time-et al 2014; Noh time-et al 2016a) As DGT devices accumulate dissolved mtime-etals, they depltime-ete the soil porewater, and weakly-bound metals are resupplied from the solid phase; thus, the application of DGT can be expanded to include the quantification of metal lability and bioavailability in soils The existing research has demonstrated that metal accumulation in aquatic plants is better predicted by DGT than by dissolved metals or leachable metals determined by the Community Bureau of Reference (BCR) and National Institute of Standards and Technology (NIST) sequential extractions (Hamels et al 2014) The successful application of DGT for metal bioavailability prediction was also shown with aquatic and benthic invertebrates (Amato et al 2014b; Wang et al 2014b; He et al 2018c) Field-based experimental results showed that DGT provides stronger predictions of Cu bioavailability than the dissolved Cu in seawater near an industrial complex

(Wang et al 2014b) Furthermore, DGT techniques have shown broader applicability to more diverse metals: the DGT-soil interface concentrations (CDGT) for Cr, Ni, Cu, and Pb were positively correlated with metal concentrations in freshwater snail tissue, while BCR extraction could only predict Pb bioavailability (Yin et al 2014a) Although multiple studies have provided substantial evidence that DGT is an effective tool for evaluating bioavailability, the applicability in this context is conditional on the type of metal and organism involved (Duquène et al 2010; Zarrouk

et al 2014; Peng et al 2017) Robust evaluations are therefore needed to use DGT as a tool for predicting the bioavailability of specific metals

Predicting the bioavailability of Hg in soils and sediments is challenging due to the high particle-water partition coefficient of Hg and active speciation in subsurface environments, such

as oxidation, reduction, methylation, and demethylation Initially, DGT was applied to estimate time-averaged Hg and monomethyl Hg (MMHg) concentrations in sediment porewater (Hong et

al 2014; Noh et al 2016a) In situ application of DGT furthermore helped to explain the biogeochemical reactions related to Hg(II) methylation by measuring porewater redox species such

as sulfide and iron (Hong et al 2014) To date, only a few studies have applied DGT to assess Hg and MMHg bioavailability for plants and invertebrates (Liu et al 2012b; Amirbahman et al 2013b) When MMHg bioavailability was monitored using a DGT approach in inland China, where rice consumption is a major source of MMHg exposure, the Hg uptake rate of the rice root was fairly well predicted using DGT-measured MMHg flux (Liu et al 2012b) Moreover, the DGT technique was able to estimate the Hg(II) methylation rate by quantifying labile MMHg concentrations in estuarine sediments (Clarisse et al 2011)

The objective of the current study is to test whether DGT can be used as a tool to evaluate the bioavailability of Hg for sedentary invertebrate in soil The earthworm Eisenia fetida was

selected for this purpose, because Eisenia fetida has a high tolerance for metals and thus it can accumulate large amounts of weakly-bound metals (Wen et al 2004) In order to obtain kinetic parameters such as Hg uptake and release rate constants by DGT or earthworms based on the time-series data, a one-compartment model (OCM) was applied, which describes Hg partitioning between the DGT probe (and earthworm) and the soil by assuming that DGT (and earthworm) is

a single and uniform compartment The uptake and release rate constants were then interpreted based on changes in organic matter content and pH to investigate the influence of soil properties

on the accumulation of Hg by DGT and earthworms Finally, Hg concentrations in earthworm tissue were compared to DGT-measured Hg flux, porewater Hg, and acid-extractable Hg concentrations to determine whether DGT has an advantage over conventional methods in predicting the bioavailability of soil Hg

2.2 Experimental methods

2.2.1 DGT probe preparation

A DGT probe has three main components: the diffusive layer, the resin layer, and a

0.45-m cellulose filter membrane The probe assembly was completed following the procedures described by Zhang &Davison 1995) The filter membranes were soaked in 2% (v/v) HNO3 for 24 hours and rinsed with deionized (DI) water (18.2 MΩ·cm) prior to use A diffusive layer with a thickness of 0.75 mm was made by dissolving 1.5 g of pure agarose in 100 mL of DI water (Yin

&Fan 2011) This agarose solution was stirred and heated sufficiently at 80°C until the solution became transparent Then, the gel was quickly cast between the prepared glass plate sets and allowed to cool and solidify at room temperature for 15 minutes The agarose gels were stored at 4°C in a 10 mM sodium nitrate solution (Warnken et al 2006b)

To prepare the resin layer, 40% acrylamide/bisacrylamide was used as a gel solution, in which the bisacrylamide works as a cross-linker (Noh et al 2015; Hong et al 2011) Subsequently,

1 g beads of 3-mercaptopropyl-functionalized silica (Sigma Aldrich) was added to a 15-mL conical tube containing 10 mL of the gel solution For polymerization, 60 L of 10% (w/v) freshly prepared ammonium persulfate and 15 L of tetramethylethylenediamine were added to the flask The gel was mixed thoroughly using a mechanical shaker before casting between glass plate sets

to make a 0.5-mm thick binding gel This layer became solid at room temperature after 45 minutes The prepared Hg-binding gels were stored in DI water (Zhang &Davison 2000) The binding gel was placed on the top of the acid-cleaned backing cylinder A diffusive gel was then placed on the binding gel Then, the 0.45-m filter membrane was positioned on the top of a diffusive gel to protect it, and then, the acid-cleaned front cover was pressed down gently to completely seal the gels

2.2.2 Preparation and characterization of soils

To explain how soil Hg concentrations affect the Hg partitioning between soil and earthworm (and DGT), four boxes of artificial soil (~1 kg each) were prepared with 69% fine sand

of 0.21 – 0.3 mm size range, 20% kaolinite, 10% peat moss, and 1% CaCO3 (Table 1), as described

in the OECD method for the standard soil preparation to test invertebrate toxicity (OECD 2016b) These were then mixed with water to achieve a 50% water content and incubated for 180 days at 20°C After 180 days of incubation, the four boxes of aged soils were spiked with HgCl2 to have nominal concentrations of 5, 50, 100, and 250 nmol Hg g-1 in dry soil To achieve 100% water content, an appropriate amount of DI water was then added to each soil box, followed by another seven days of incubation

To identify how total organic carbon (TOC) in soil affects the Hg distribution in soil, earthworm, and DGT, four further boxes of artificial soil (~1 kg each) were prepared with 5, 10,

15, and 20% peat moss, 59–74% fine sand, 20% kaolinite, and 1% CaCO3 (Table 1) After seven days of incubation at 20°C with 50% water content, these fresh soils were spiked with an HgCl2

solution to have a nominal concentration of 50 nmol Hg g-1 in dry soil To achieve the same water level for all the soil boxes, appropriate amounts of DI water were added, followed by another seven days of incubation at room temperature The same preparation process was used to test the soil pH effect on the Hg distribution between soil, earthworm, and DGT The soil pH was controlled by adjusting the CaCO3 content to be 0.1, 0.2, and 0.5% Similar to the peat moss experiment, the soils were incubated for seven days with 50% water content After spiking with an HgCl2 solution

to obtain a nominal Hg concentration of 50 nmolg-1 in dry soil, 100% water content was obtained

by adding extra DI water and the soils were incubated for another seven days

Both the 𝑝𝐻 and 𝑝𝐻 were determined using OECD methods (OECD 2016a) Briefly, a mixture of soil and water (or 0.01 M CaCl2 solution) was prepared at a ratio of 1:5 (w/v) and left to settle for two hours The pH of the liquid was measured using a pH meter probe (Orion

5 Star, Thermo Scientific) The soil TOC was determined using a Euro-EA elemental analyzer The maximum water holding capacity (MWHC) of each soil was determined as described in the OECD guidelines: 5 g of dry soil was prepared in a plastic tube and the bottom was covered with

a filter paper The soils were submerged in water until saturated for two hours, and the tubes were then placed on wet sand to allow excess water to be drawn out by gravity The MWHC of the soil was determined as the amount of water held per dry weight of the sample in % unit

Table 2.1 Composition and chemical characteristics of artificial soils for diffusive gradients in thin films (DGT) and

earthworm deployment: pH, total organic carbon (TOC), aging period, water content, and maximum water holding capacity (MWHC)

Soil Type Sand (%) Kaolinite (%)

Peat moss (%)

Aging period (days)

Water content (%)

MWHC (%)

2.2.3 Deployment of DGT probes and earthworms

The assembled DGT units, stored in DI water to prevent the diffusive layer and resin gel from drying out, were deployed in each soil box For proper arrangement, a small amount of soil was used to remove any air pockets between the DGT probe and the soil surface Then, the DGT units were placed on the surface of the artificial soils in triplicate The DGT probes were recollected at 0.5, 1, 2, 5, and 10 days Earthworms (E fetida) were purchased from Caroline Biological Supply Co., USA They were cultured with defined conditions, and 80% humidity and

a 12:12 hour light:dark cycle were maintained during the culturing period Around 120–150 adult earthworms of consistent size (0.4–0.5 g) were selected for deployment in soils After earthworms were collected from the incubated soils, they were rinsed with DI water to remove soil particles Then the earthworms were depurated on wet filter paper in petri dishes without being fed The petri dishes were left in an incubated cabinet with a constant temperature of 20ºC with a 12:12 hour light:dark cycle for 24 hours Approximately 35-40 depurated earthworms were deployed in each soil box, and six worms were retrieved at 0.5, 1, 2, 5, and 10 days

2.2.4 Hg measurement in DGT probes, earthworms, and soils

The resins collected from the DGT probes were rinsed with DI water and stored at 4°C They were cut into small pieces so that the mass was appropriate for the Hg detection range of a direct Hg analyzer (DMA) A working solution diluted from 1,000 ppm Hg standard with 5% nitric acid (Fluka, USA) was used to calibrate the DMA The marine sediment MESS-3 (0.091 ± 0.009 ppm, National Research Council of Canada) was used as a certified reference material (CRM), and the CRM recovery averaged 103 ± 5% (n=10) The optimal conditions for direct resin measurement without elution were as follows: drying at 200°C for 70 sec, increasing the temperature to 650°C for 90 sec, and then decomposition at 650°C for 90 sec Mercury

concentration is blank resin was found to be 1.4 ± 0.7 pmol cm-2 (n=5) When Hg uptake efficiency, after 12 hours of resin deployment in DI water, was calculated as [(Hg in resin)/(Hg in DI waterinitial)]×100, it was 91 ± 3.4% (0.81-0.95 nmol cm-2, n=8) The recovery of Hg after 12 hours

of resin deployment in DI water was estimated as (Hg in resin)/[(Hg in DI water)initial - (Hg in DI water)final]×100, and it was 99.7% (n=8)

Approximately 5 g of soil was collected from the surface (0 – 2 cm depth) of each box to

a 15 mL conical tube after retrieving DGT probes at each time interval (0.5, 1, 2, 5, and 10 days) The tubes were then centrifuged at 3,000 rpm for 30 min to separate soil solution and particles Porewater was then collected by filtering the soil solution through a 0.45-m PEFE filter membrane (Whatman) and preserved with 0.4% v/v hydrochloric acid for Hg measurements, meanwhile soil particles were freeze-dried and stored at 4oC for later analysis Similar to the resin

Hg, Hg concentrations in freeze-dried soils were measured using the DMA, and the measurement conditions were the same as for the DGT resins The optimal DMA conditions for analyzing the porewater Hg were as follows: drying at 200°C for 90 sec, increasing the temperature to 650°C for

90 sec, and then final decomposition at 650°C for 150 sec The CRM (MESS-3) recovery for soils averaged 101 ± 8% (n=20) Labile Hg concentrations in the freeze-dried soils were determined using 0.2 M nitric acid (Fernández-Martínez &Rucandio 2013) To do so, 20 mL of 0.2 M HNO3

was added to 0.5 g of dried soil, and this soil solution was mixed using a rotary shaker for two hours to release labile Hg from the soil The extracted solution was centrifuged and filtered through

a 0.45-m filter prior to Hg analysis

The earthworms retrieved at each time interval were rinsed with DI water to remove soil and left to depurate on wet filter paper in a petri dish for 48 hours The clean earthworms were then freeze-dried for 24 hours before total Hg was analyzed by DMA The optimal DMA

conditions for earthworm analysis were found to be as follows: drying at 200°C for 90 sec, increasing the temperature to 650°C for 90 sec, and then final decomposition at 650˚C for 90 sec The CRM recovery for the earthworms averaged 103 ± 5% (n=10)

2.2.5 Calculation of CDGT and OCM simulation

The time-averaged Hg concentration at the DGT-soil interface (CDGT) is calculated by equation (2.1):

C = M∆g

DAt (2.1) where M is the accumulated Hg in the resin (nmol), ∆g is the thickness of the diffusive gel (cm), D is the diffusion coefficient (cm2 sec-1), A is the exposure area (cm2), and t is the exposure time (sec) An OCM was used to estimate the Hg uptake and elimination rate constants by the earthworms and the DGT resins The OCM describes the process of metal distribution and elimination in the DGT resin or earthworm body, by assuming that the entire resin or body acts like a single and uniform compartment (Bade et al 2012) Once metal enters to the compartment,

it distributes and equilibrates rapidly throughout the compartment, and metal elimination from the compartment begins immediately after its entering Fitting the model to the experimental data allows to calculate the uptake rate constant (k1) and elimination rate constant (k2), as defined by the equation (2.2):

dC

dt = k , C − k , C (2.2) where Cs is the Hg concentration in the soil compartment (nmol g-1), Cx is the mass of Hg

in DGT resin (nmol cm-2) or earthworm (nmol g-1), k1,x is the uptake rate constant for DGT resin (g cm-2 day-1) or earthworm (g g-1 day-1), and k2,x is the elimination rate constant (day-1) for DGT

resin or earthworm The integration of equation 2 yields the solution (2.3):

C (t) = C (0)e , +k , C

k , 1 − e , (2.3) where Cx(0) is the initial mass of Hg in DGT resin (nmol cm-2) or earthworm (nmol g-1) The uptake and elimination rate constants (k1,x and k2,x) can be determined from the best fit to the accumulated mass of Hg versus time The steady-state refers to the situation when Hg accumulation is in equilibrium with its elimination, and at steady-state (tss), equation (3) becomes the following:

C (t) =k , C

k , (2.4) The DGT-soil accumulation factors (DSAF) and earthworm-soil accumulation factor (BSAF) at steady-state are defined as follows:

J = C k , (2.7) 2.3 Results and discussion

2.3.1 Effect of Hg concentration on earthworm and DGT accumulation of Hg

The accumulated Hg ranged from 0.086 to 0.32 nmol cm-2 for DGT resin and from 5.7 to

15 nmol g-1 for earthworms on day 10, when DGT probes and earthworms were deployed in the soils with measured total Hg concentration of 5.2, 56, 105, and 262 nmol g-1 (Fig 2.1) The highest

Hg accumulation was shown for the highest Hg soils, with a constant Hg concentration after 5

days for DGT and 2 days for earthworms In the lowest Hg treatment, a continuous increase of Hg was observed for resins until day 10; on the contrary, Hg concentration in earthworm was constant after day 2

Figure 2.1 The accumulation of Hg in diffusive gradients in thin films (DGT) resin and earthworm tissue as a function of deployment time for aged soils The experimental Hg data were fitted by one compartment model (OCM) Soil composition was 69% sand, 20% kaolinite, 10% peat moss, and 1% CaCO3, and the measured Hg concentration was A 5.2 nmol g-1, B 56 nmol g-

1, C 105 nmol g-1, and D 262 nmol g-1

The experimental and model based DSAF decreased with an increase in soil Hg concentration (Table 2) The Hg species could diffuse rapidly through the diffusive gel and bind strongly to thiols on the resin gel at low soil Hg concentrations When soil Hg concentrations are greater, the strongest binding sites on the DGT sorbent are more likely to become saturated and only the weak binding sites would eventually be available for binding, as was implied by the increased elimination rate constant with an increase of soil Hg concentration (Table 2) The same trend was found for the experimental and modeled BSAF: the BSAF decreased with an increase

in soil Hg concentration as the Hg elimination rate constant for earthworm increased along with

an increase of soil Hg concentrations, with the exception of the 5 nmol g-1 Hg These initial results suggest that DGT could be useful for predicting bioaccumulation of labile Hg in soil

Table 2.2 Output parameters of the one compartment model [i.e., diffusive gradients in thin films (DGT) and earthworm uptake rate constant, elimination rate constant, DGT-soil accumulation factor (DSAF), and biota-soil accumulation factor (BSAF)],

and experimentally determined DSAF and BSAF

Soil type

Uptake rate constant (gcm-2 d-

1)

Eliminati

on rate constant (d-1)

Model DSAF (gcm-2)

Exp DSAF (gcm-2)

Uptake rate constant (gg-1 d-1)

Eliminati

on rate constant (d-1)

Model BSAF (gg-1)

Exp BSAF (gg-1)

2.3.2 Effect of peat moss content on earthworm and DGT accumulation of Hg

The DGT and earthworm Hg accumulations for peat moss contents of 5, 10, 15, and 20%, corresponding to 1.6, 2.2, 3.8, and 4.8% TOC, respectively, are shown in figure 2.2 The accumulated mass of Hg ranged from 0.25 to 1.68 nmol cm-2 for DGT resin and from 43 to 356 nmol g-1 for earthworms at day 10 In comparison to the aged soils used for Hg concentration tests, these fresh soils showed an earlier steady-state within 2–4 days for both DGT and earthworms Higher Hg concentration at steady-state was found by both measurements (DGT and earthworm)

in soils with lower concentrations of peat moss The Hg uptake rate constant, k1, decreased from 2.58×10-2 to 8.22×10-3 gcm-2 d-1 for DGT and from 4.50 to 0.43 gsoil g-1

EW d-1 for earthworms as the peat moss content increased from 5 to 20% (Table 2) On the contrary, the elimination rate constant, k2, was relatively stable, and consequently simulated DSAF and BSAF decreased as the peat moss content increased (Fig S2.1)