Use of non-linear multiple regression models for setting water quality criteria for copper: Consider the effects of salinity and dissolved organic carbon

Abstract

Cu pollution is a global concern because of its high toxicity and persistence. Few investigations have been conducted on the effects of salinity and dissolved organic carbon (DOC) on Cu toxicity and water quality criteria (WQC). To analyze their impact on the WQC of Cu, non-linear multiple regression (NLMR) models based on salinity and DOC were constructed. The NLMR models demonstrated that when salinity increased, the toxicity values for Cu on fish, mollusca, rotifer, and echinodermata first increased and then declined, whereas those for arthropoda and algae increased. These findings demonstrate that salinity has a substantial impact on Cu toxicity, primarily owing to changes in physiological activity. The original and corrected WQC values in the upper, middle, and outer regions of the Yangtze Estuary were derived based on the species sensitivity distribution method. These values were 1.49, 3.49, 8.86, and 0.87 μg·L−1. An important finding was that lower Cu concentrations in the outer areas posed the highest ecological risk owing to the effects of salinity and DOC. NLMR models are applicable to other coastal areas worldwide. This provides valuable information for the establishment of an accurate and protective estuary for Cu-related WQC.