The effects of organic matter, pH, and dissolved ligands on the mobility of cadmium in soils : A thesis submitted in partial fulfilment of the requirements for the Degree of Doctor of Philosophy at Lincoln University

The fate of cadmium (Cd) in soil is important because of its toxicity even at low concentrations. Mobilisation of Cd to soil below the root zone can decrease Cd accumulation in topsoil where plants take up available Cd. However, this mobilisation may ultimately result in groundwater contamination. The majority of Cd in soil usually occurs in the solid phase, with only small amounts present in soil solution, which are the most mobile. The partitioning of Cd between matrix and soil solution is affected by soil pH, organic matter content, and dissolved ligands in porewater. The aim of this study was to investigate how these factors may affect the potential mobility of Cd in soils. Experiment (1): Ten soil amendments that are rich in organic matter (OM), but with different provenances, were tested for their ability to sorb Cd across a pH range of 4.5 to 6.5. These amendments included five different composts, two peats, two biosolids and spent coffee grounds. By using an adapted diffusive gradient thin-film (DGT) method, the lability of Cd associated with dissolved organic matter (DOM) emergent from two of the above composts and two peats were tested and related to the quality of DOM. The retention of Cd varied between the OM amendments and was generally reflected by the cation exchange capacity (CEC) of each amendment. The Cd sorption capacity increased with pH, implying that variable charge sites play an important role in Cd sorption to OM. The Cd-DOM species emergent from the composts appeared to be more kinetically limited than that from the peats and the aromaticity of DOM was a less important factor in determining this lability of Cd-DOM complexes. Experiment (2): To explore the ability of OM to immobilise the soil Cd, a compost and a peat which showed high and moderate Cd sorption capacities, respectively were amended with three different soils spiked to about 2 mg kg-1 Cd at three pHs (5.6, 6.4 and 7.2). A short-term column leaching experiment was carried out to investigate the potential mobility of Cd in these OM amended soils and how that was related to quality and quantity of DOM leached. The “truly” dissolved (<5 kDa) Cd fraction from the mobilised dissolved Cd (<0.45 µm) was separated by ultrafiltration. The quality of DOM mobilised was assessed with ultraviolet absorbance measured at various wavelengths and fluorescence excitation-emission matrices. At relatively low pH (5.6) the compost was more effective in immobilising Cd than the peat, which consistent with the first sorption experiment. However, regardless of the high capability of soil solid phases (including OM amended) to bind Cd, mobilisation at higher pH (6.4) was linked to DOM mobilisation. Approximately 50% of Cd mobilised from OM amended soils existed in the dissolved colloidal fraction (5 kDa-0.45 µm). This mobile colloidal-Cd was highly correlated with the mobilised DOM (r=0.77, p=1.74 X 10-11), in which Cd binding ligands in non-aromatic structures in the high molecular weight DOM may have played a major role in Cd complexation. The mobilised DOM from the compost and the peat amended soils had a greater humification index than the DOM mobilised from the soil itself and, therefore mobile Cd-DOM from those OM amended soils may persist in soil for a longer period. Cadmium mobility was also affected by physicochemical differences between soils. Sorption of Cd or OM to mineral surfaces in an allophanic clay-rich soil may have decreased Cd-DOM mobility compared to more sandy soils. Experiment (3): Investigated potential soil Cd mobilisation from soils treated with cow urine, also in a column leaching experiment, employing different three soils to before. This experiment revealed the potential mobility of Cd species in soil solution and Cd released to soil solution from readily exchangeable sites. Here, top 1/3 of columns was packed with Cd isotopes (108Cd and 116Cd) spiked soil while the rest of 2/3 was unspiked soil. Cadmium isotope ratio (116/108Cd) was used to investigate the extent of mobilisation of Cd-originated from the spiked soil and Cd speciation in the leachates. Overall, 17 to 60% of total Cd leached originated from the top spiked soil layer across three soils, which did not receive urine. The fractional contribution of Cd from spiked soil for total Cd leached increased with pH. Cow urine increased dissolved chloride (Cl-) and DOM in soils. While 7-10% of Cd mobilised was as Cd-chloro complexes, Cl- concentration alone could not explain Cd mobilisation. This was attributed to the higher affinity of Cd for colloidal DOM, where mobilised Cd from urine treated soils had a strong positive correlation with the mobilised high molecular weight DOM (humic acid-like: r=0.73, p=7.2 X 10-7). Consistent with the first leaching experiment, more Cd mobilised from coarser soils than from soils with low sand content. Experiment (4): The size of labile Cd pool and resupply kinetics for three stable Cd isotopes (108Cd,112Cd and 116Cd) in the soils from the third experiment was explored through a combination of experimental measurements using DGT and numerical modelling the DGT-Induced Fluxes in Soils model (DIFS). The labile Cd pool size for 108Cd and 116Cd isotopes were similar but larger than for 112Cd within soils, which appeared to be affected by the contact time of each isotope. Desorption kinetics of 108Cd and 116Cd isotopes in same soil were similar. The labile pool size and resupply kinetics for each isotope were different between soils. The labile pools of the three Cd isotopes were relatively smaller in the allophanic clay-rich soil, which also had a greater sand content than the other two soil. This study found that the amount of Cd transported in OM amended soils depends on the type of the OM used, and the extent to which this happens is controlled by the interaction between the type of OM, soil type, and its pH. Further, the relative contribution of Cd2+, Cd-DOM and Cd-chloro species towards Cd transport depends on properties of soil, pH, OM content, DOM and Cl- concentrations. Finally, this work provides new insights for the use of stable Cd isotopes in tracing the transport of topsoil Cd to subsoil and/or groundwater.