Premium
Development of a regression model to predict copper toxicity to Daphnia magna and site‐specific copper criteria across multiple surface‐water drainages in an arid landscape
Author(s) -
Fulton Barry A.,
Meyer Joseph S.
Publication year - 2014
Publication title -
environmental toxicology and chemistry
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.1
H-Index - 171
eISSN - 1552-8618
pISSN - 0730-7268
DOI - 10.1002/etc.2631
Subject(s) - alkalinity , biotic ligand model , environmental science , surface water , dissolved organic carbon , environmental chemistry , daphnia magna , perennial stream , hydrology (agriculture) , chemistry , ecology , streams , toxicity , environmental engineering , biology , geology , computer network , geotechnical engineering , organic chemistry , computer science
The water effect ratio (WER) procedure developed by the US Environmental Protection Agency is commonly used to derive site‐specific criteria for point‐source metal discharges into perennial waters. However, experience is limited with this method in the ephemeral and intermittent systems typical of arid climates. The present study presents a regression model to develop WER‐based site‐specific criteria for a network of ephemeral and intermittent streams influenced by nonpoint sources of Cu in the southwestern United States. Acute (48‐h) Cu toxicity tests were performed concurrently with Daphnia magna in site water samples and hardness‐matched laboratory waters. Median effect concentrations (EC50s) for Cu in site water samples ( n = 17) varied by more than 12‐fold, and the range of calculated WER values was similar. Statistically significant (α = 0.05) univariate predictors of site‐specific Cu toxicity included (in sequence of decreasing significance) dissolved organic carbon (DOC), hardness/alkalinity ratio, alkalinity, K, and total dissolved solids. A multiple‐regression model developed from a combination of DOC and alkalinity explained 85% of the toxicity variability in site water samples, providing a strong predictive tool that can be used in the WER framework when site‐specific criteria values are derived. The biotic ligand model (BLM) underpredicted toxicity in site waters by more than 2‐fold. Adjustments to the default BLM parameters improved the model's performance but did not provide a better predictive tool compared with the regression model developed from DOC and alkalinity. Environ Toxicol Chem 2014;33:1865–1873 . © 2014 SETAC