Influence of soil solution properties and predicting model on soil soluble copper toxicity to tomato shoot

Abstract: Metal bioavailability and toxicity is not only determined by its speciation and contents, but also by the soil water chemistry. In the present study, bioassays of tomato shoot were performed in 17 Chinese soils to evaluate the phytotoxicity of soluble copper (Cu) toxicity based on soil pore water and 0.01 M CaCl2 extraction. The selected soils, which represented the major soil types and properties in China, were spiked with soluble Cu chloride. Sub-samples of the spiked soil were leached by artificial rainwater to simulate the field conditions and the toxicity was compared with that in unleached soils. When considering the toxicity thresholds for Cu in soil pore water, it was found that the effective concentrations that caused 10% root growth inhibition (EC10) and 50% inhibition (EC50) varied widely from 0.06 to 1.47 mg/L and from 0.17 to 3.42 mg/L in 17 unleached soils and from 0.05 to 2.24 mg/L, from 0.13 to 4.37 mg/L in leached soils, representing 23 to 41.1 folds differences. Similarly, the toxicity thresholds for Cu extracted by 0.01 M CaCl2 showed that the EC10 and EC50 values for unleached or leached soils also varied considerably from 0.18 to 2.64 mg/L and from 0.57 to 6.14 mg/L in unleached soils and from 0.18 to 1.28 mg/L, from 0.61 to 7.11 mg/L in leached soils, representing 6.9 to 14.4 fold differences. However, leaching did not significantly decrease soluble Cu toxicity thresholds in most soils. These results indicated that the toxicity thresholds of soluble Cu variations were influenced by soil solution properties in a wide range of soils. Meanwhile, the relationships were developed between soil solution properties and phytotoxicity threshold values for copper in a wide range of soils. The multiple regression results showed that Ca2+ and dissolved organic carbon (DOC) were the two most important factors affecting the extent of Cu toxicity in soil pore water on tomato shoot in leached and unleached soils, respectively, and meanwhile they were positively related to the toxicity thresholds. Single Ca2+ was found to explain 64% and 31% of the variance in soluble Cu toxicity threshold EC10 and EC50, respectively, in pore water across unleached soils. For leached soils, single DOC was found to explain 48% and 56% of the variance of EC10 and EC50. Soil solution pH was not the most important factor controlling the soluble Cu toxicity, while it could improve the prediction of the model to some extent, with the toxicity thresholds in pore water increasing as pH increased. When incorporating these parameters (DOC, pH, electrical conductivity (EC), S, Ca2+, Mg2+, K+ and Na+) into the regression models, together with corresponding toxicity thresholds for soluble Cu, the coefficient of determination (r2) for EC10 or EC50 ranged from 0.75 to 0.99 in leached and unleached soil. This implied that the soluble Cu toxicity on tomato shoot could be better estimated by soil pore water chemistry. These quantitative relationships between soluble Cu toxicity and soil solution properties can contribute to the development of a soluble Cu toxicity risk assessment of the terrestrial environment in China.