The objective of this study was to determine whether the water effect ratio (WER) or biotic ligand model (BLM) could be applied to efficiently develop water quality criteria (WQC) in Korea. Samples were collected from 12 specific sites along the Yeongsan River (YSR), Korea, including two sewage treatment plants and one estuary lake. A copper toxicity test using Daphnia magna was performed to determine the WER and to compare to the BLM prediction. The results of the WER from YSR samples also indicated significantly different copper toxicities in all sites. The model-based predictions showed that effluent and estuary waters had significantly different properties in regard to their ability to be used to investigate water characteristics and copper toxicity. It was supposed that the slight water characteristics changes, such as pH, DOC, hardness, conductivity, among others, influence copper toxicity, and these variable effects on copper toxicity interacted with the water composition. The 38% prediction was outside of the validation range by a factor of two in all sites, showing a poor predictive ability, especially in STPs and streams adjacent to the estuary, while the measured toxicity was more stable. The samples that ranged from pH 7.3-7.7 generated stable predictions, while other samples, including those with lower and the higher pH values, led to more unstable predictions. The results also showed that the toxicity of Cu in sample waters to D. magna was closely proportional to the amounts of acidity, including the carboxylic and phenolic groups, as well as the DOC concentrations. Consequently, the acceptable prediction of metal toxicity in various water samples needs the site-specific results considering the water characteristics such as pH and DOC properties particularly in STPs and estuary regions.
Keywords: Biotic ligand model; Copper; Daphnia magna; Toxicity; Water-effect ratio.
Copyright © 2017 Elsevier Inc. All rights reserved.