Improving Sediment Toxicity Assessment by Considering Iron Oxide?Induced Cadmium Bioavailability in Oxidized Sediments = 산화 퇴적토에서 철 산화물에 의한 카드뮴의 생물학적 이용성 변화를 고려한 독성 평가 방법 개선

국문 초록 (Abstract)

퇴적물의 독성 평가를 정확하게 수행하는 것은 효과적인 환경 관리를 위해 필수적이지만, 퇴적물 구성, 오염물질의 분배, 생물학적 이용성 간의 복잡한 상호작용으로 인해 여전히 어려움이 있다. Sediment Effect Concentrations (SECs)와 같은 전통적인 퇴적물 관리 지침은 오염물질 농도를 기준으로 독성을 분류하지만, 생물학적 이용성을 충분히 고려하지 않아 경계값 근처의 시료에서 예측 정확도가 낮다는 한계가 있다. Equilibrium partitioning (EqP) 기반 모델은 이러한 문제를 일부 보완했지만, 일반적으로 유기탄소와 산휘발성 황(AVS)만을 고려하기 때문에 산화 환경에서 주요 흡착상인 철 산화물의 기여는 반영하지 못한다.
본 연구는 담수 퇴적토 내 카드뮴(Cd)의 독성 평가를 위해 다양한 흡착상과 생물학적 이용성 지표를 통합한 평가 프레임워크를 개발하고 검증하였다. 먼저, pH, 총 유기탄소(TOC), 그리고 비정질 및 결정질 철 산화물을 입력 변수로 활용한 Cd의 총 분배계수(Kd) 예측 모델을 개발하였다. WHAM 7 모델 시뮬레이션 및 흡착 실험 결과, pH에 따라 Kd 값이 최대 세 자릿수 이상 차이를 보였으며, 특히 비정질 철 산화물은 pH 8에서 367.4 L/g의 가장 높은 흡착력을 나타냈다. 본 연구를 통해 제안한 모델은 전국 21개 실제 퇴적물 시료에 적용했을 때 측정값 대비 10배 이내의 정확도로 총 Kd를 예측 하였다.
두 번째로, 이러한 방식으로 예측한 Kd를 바탕으로 공극수 내 Cd 농도(Cw)를 계산하고, 이를 만성 독성 기준인 Final Chronic Value (FCV = 0.72 µg/L)로 정규화하여 Interstitial Water Toxic Unit (IWTU)를 산정하였다. Hyalella azteca를 이용한 생물 독성 실험 시험 결과와 비교한 결과, 기존 SEC만을 활용했을 때 43%였던 예측 정확도가 통합 프레임워크 적용 시 76%로, 민감도는 36%에서 79%로 향상되었다. 따라서 본 연구에서는 이를 바탕으로 SEC와 IWTU를 병합한 단계적(tiered) 독성 평가 체계를 제안하였다.
마지막으로, 산화 환경에서 Cd의 시계열 거동을 규명하기 위해 Cd가 주입된 퇴적토를 168일간 수로(flume) 시스템에서 산화시켜 실험을 진행하였다. 실험 결과, 시간이 지남에 따라 비정질 철 산화물 농도가 유의미하게 증가하였고, 이는 Cd의 고형상 흡착을 촉진하여 Cw 및 IWTU가 감소하는 결과를 가져왔다. 예를 들어, 고농도 Cd로 오염시킨 표층 퇴적토 시료에서 Cw는 33.4 µg/L에서 6.3 µg/L로, IWTU는 46.4에서 8.8로 감소하였다. 그러나 IWTU 값은 여전히 독성 기준을 초과하였고, H. azteca는 전 시료에서 100% 사망하였다. 반면, AVS와 TOC 기반의 s-BLM은 대부분의 시료에서 비독성으로 예측하여, 실제 생물 독성 실험 결과와 큰 차이를 보임으로써 산화 환경에서의 적용에는 한계가 있음이 드러났다.
본 연구는 전통적인 AVS 기반 모델의 한계를 지적하고, 산화 조건에서 철 산화물 및 시계열 퇴적토 조성 변화를 반영한 독성 평가의 중요성을 강조한다. 제안된 통합 프레임워크는 산화성 퇴적물에 대한 보다 현실적이고 신뢰도 높은 독성 평가를 가능하게 하며, 향후 환경 위해성 평가 및 관리에 활용될 수 있을 것으로 기대된다.

다국어 초록 (Multilingual Abstract)

Accurate assessment of sediment toxicity is essential for effective environmental management but remains challenging due to the complex interactions among sediment composition, contaminant partitioning, and bioavailability. Traditional sediment quality guidelines (SQGs), such as Sediment Effect Concentrations (SECs), classify sediments based on contaminant concentrations but often fail to account for bioavailability, particularly in borderline cases. While models based on the equilibrium partitioning (EqP) theory have improved predictive capabilities, they typically focus only on organic carbon and acid volatile sulfides (AVS), limiting their applicability under oxidizing conditions where other sorption phases, such as iron oxides, play a dominant role in metal binding.
This dissertation develops and validates a comprehensive framework for cadmium (Cd) toxicity assessment in freshwater sediments by incorporating multiple sorption phases and bioavailability indicators. First, a mechanistic prediction model for the sediment–water partition coefficient (Kd) of Cd was developed based on pH, total organic carbon (TOC), and amorphous/crystalline iron oxides. WHAM 7 simulations and adsorption experiments revealed that Kd values vary by over three orders of magnitude across pH ranges, with amorphous iron oxides showing the highest sorption affinity (up to 367.4 L/g at pH 8.0). The model successfully estimated Kd values for 21 field sediments within a 10-fold range of measured values, outperforming previous organic carbon–based predictions.
Second, using the predicted Kd values, porewater Cd concentrations (Cw) were calculated and normalized to the Final Chronic Value (FCV = 0.72 µg/L) to derive Interstitial Water Toxic Units (IWTU). When compared against survival data from Hyalella azteca bioassays, this bioavailability-based approach significantly improved the classification accuracy of SECs, increasing overall predictability from 43% to 76% and sensitivity from 36% to 79%. A tiered toxicity assessment framework was proposed by integrating SEC thresholds with IWTU evaluations, enhancing reliability, especially for samples previously classified as “Uncertain.”
Lastly, to investigate the temporal dynamics of Cd behavior under oxidizing conditions, a 168-day flume experiment was conducted with Cd-spiked sediments. Sediment geochemistry showed a significant increase in amorphous Fe oxides over time, which facilitated Cd immobilization and led to a steady decline in predicted Cw and IWTU values. For instance, in surface sediment of the high-Cd treatment, Cw decreased from 33.4 µg/L to 6.3 µg/L, while IWTU dropped accordingly from 46.4 to 8.8. However, despite this decline, all IWTU values remained above toxic thresholds, aligning with 100% mortality observed in bioassays. In contrast, the sediment biotic ligand model (s-BLM), which relies on AVS and TOC, failed to predict toxicity in most cases, underestimating risk due to the persistence of non-reactive AVS under oxic conditions.
Overall, this research underscores the limitations of conventional AVS-based models and highlights the importance of including iron oxides and temporal geochemical changes in toxicity assessments. The proposed bioavailability-oriented, tiered framework offers a robust and adaptable tool for sediment quality evaluation in oxidizing environments.

목차 (Table of Contents)

• Abstract i
• Table of contents iv
• List of Tables x
• List of Figures xii
• Chapter 1.
• Introduction 1
• 1.1. General overview 1
• 1.2. Background 3
• 1.2.1. Sediment effect concentration (SEC) 3
• 1.2.2. Equilibrium partitioning approach 6
• 1.2.3. Metal adsorption on Fe oxides 9
• 1.2.4. Use of Hyalella azteca in sediment toxicity assessment 10
• 1.3. Objective and scope 12
• 1.4. Organization of dissertation 13
• References 15
• Chapter 2.
• Enhanced Prediction of Cadmium Partitioning in Oxidized Sediments 20
• 2.1. Introduction 20
• 2.2. Materials and methods 24
• 2.2.1. Preparation of sediment samples 24
• 2.2.2. Analysis of physicochemical properties of sediment 24
• 2.2.3. Cd adsorption on sediment samples 26
• 2.2.4. Selective extraction method 27
• 2.2.5. Kd values of humic acid and Fe(OH)3 obtained from WHAM 7 29
• 2.2.6. Kd values of ferrihydrite 31
• 2.2.6.1 Synthesis of ferrihydrite 31
• 2.2.6.2 Cd adsorption on ferrihydrite 32
• 2.2.7. Derivation of total Kd value 32
• 2.2.8. Statistical analysis 35
• 2.3. Results and discussion 36
• 2.3.1. Physicochemical characteristics of sediments 36
• 2.3.2. Contribution of Fe oxides in sediment samples 40
• 2.3.3. Plotting of Kd values across a wide pH range 41
• 2.3.3.1. Kd values obtained by WHAM 7 for humic acid and Fe(OH)3 41
• 2.3.3.2. Experimentally determined Kd values for ferrihydrite 43
• 2.3.4. Comparison of predicted and measured total Kd 47
• 2.3.5. Influence of sediment components on total Kd: a multivariate approach 50
• 2.3.6. Sensitivity of the predicted Kd values 55
• 2.4. Conclusions 58
• References 61
• Chapter 3.
• Improving Predictive Performance of Sediment Toxicity Assessments by Integrating Bioavailability Metrics with Sediment Effect Concentrations 69
• 3.1. Introduction 69
• 3.2. Materials and methods 72
• 3.2.1. Use of sediment samples for SEC evaluation 72
• 3.2.2. Determination of SECs 74
• 3.2.3. Determination of aqueous Cd concentration from the predicted total Kd 75
• 3.2.4. Calculation of interstitial water toxic units (IWTU) 76
• 3.2.5. Toxicity test using Hyalella azteca 77
• 3.2.6. Quality assurance and quality control (QA/QC) 79
• 3.2.7. Statistical analysis 80
• 3.3. Results and discussion 80
• 3.3.1. Variations of physicochemical properties in sediment samples 80
• 3.3.2. SECs derived from sediment data 83
• 3.3.3. Enhanced prediction accuracy using SECs and Kd-based approaches 85
• 3.3.4. Relationship between predicted aqueous Cd concentrations and bioassay results 89
• 3.3.5. Enhanced predictability through integration of SECs and IWTU approaches 91
• 3.3.6. Proposed framework for integrated toxicity assessment 97
• 3.4. Conclusions 101
• References 104
• Chapter 4.
• Temporal Changes in Cadmium Partitioning in Sediments Undergoing Oxidation: Role of Iron Oxide Formation in a Flow-Through System 110
• 4.1. Introduction 110
• 4.2. Materials and methods 113
• 4.2.1. Preparation of sediment samples 113
• 4.2.2. Specification of the flume 114
• 4.2.3. Sediment sampling 115
• 4.2.4. Measure of sediment characteristics 116
• 4.2.5. Toxicity test using Hyalella azteca 118
• 4.2.6. Statistical analysis 119
• 4.2.7. Estimation of predicted toxicity 120
• 4.3. Results and discussion 121
• 4.3.1. Initial sediment characteristics prior to oxidation 121
• 4.3.2. Changes in sediment characteristics during oxidation 124
• 4.3.3. Cd partitioning behavior during sediment oxidation 129
• 4.3.4. Comparison of measured and predicted porewater Cd concentrations (Cw) 134
• 4.3.5. Temporal variation in predicted porewater Cd concentrations (Cw) 136
• 4.3.6. Interstitial water toxicity estimation based on predicted Cw 138
• 4.3.7. Comparison of predicted toxicity (IWTU and s-BLM) with observed mortality 140
• 4.4. Conclusions 144
• References 147
• Chapter 5.
• Conclusions and Future works 153
• 5.1. Conclusions 153
• 5.2. Future works 155
• Abstract in Korean 158