Multiple Linear Regression Modeling Predicts the Effects of Surface Water Chemistry on Acute Vanadium Toxicity to Model Freshwater Organisms

Abstract

Multiple linear regression (MLR) modeling has been successfully used to predict how water chemistry variables influence the toxicity of cationic metals to aquatic organisms, but no MLR model exists for vanadium (V). Recent research has indicated that an increase in pH (from 6 to 9), or high concentrations of sodium (473 mg Na+/L), increase V toxicity to Daphnia pulex. In contrast, increases in alkalinity (>100 mg as CaCO3) and sulfate (>100 mg SO42–/L) reduce V toxicity. How these variables influence V toxicity to Oncorhynchus mykiss (rainbow trout) was still unknown. Our results show that increasing pH from 6.2 to 8.9 tended to decrease the 96‐h median lethal concentration (LC50) for V toxicity to O. mykiss by 9.6 mg V/L. An alkalinity increase from 71 to 330 mg/L as CaCO3 tended to increase the 96‐h LC50 by 3.3 mg V/L, whereas when SO42− rose from 150 to 250 mg/L, the LC50 significantly increased by 0.3 mg V/L followed by a significant decrease of 1 mg V/L when SO42− was >250 mg/L. Sodium (between 100 and 336 mg/L) showed no effect on V toxicity to O. mykiss. The toxicity patterns for O. mykiss were similar to those observed for D. pulex, except for that of SO42−, potentially indicating different mechanisms of V uptake or regulation in the 2 species. The LC50s and associated water chemistry were combined to develop an MLR model for O. mykiss and D. pulex. Alkalinity and pH modified V toxicity to both species, whereas SO42− influenced V toxicity to D. pulex. Overall, MLR models should be considered for creating new local benchmarks or water quality guidelines for V. Environ Toxicol Chem 2020;39:1737–1745. © 2020 SETAC

Saccharide Concentration Prediction from Proxy Ocean Samples Analyzed Via Infrared Spectroscopy and Quantitative Machine Learning

“…50 Absorbance FTIR spectra generally follow a linear relationship of intensity with respect to concentration, which is advantageous for determining the new sample composition. Recent work has utilized multiple linear regression to identify heavy metals, including investigating the effect of surface chemistry on vanadium 51 and lead 52 toxicity. However, the simplicity of the method ultimately restricts the model's usefulness in more complex, dynamic systems.…”

Saccharide concentration prediction from proxy sea surface microlayer samples analyzed via infrared spectroscopy and quantitative machine learning

“…46 Absorbance FTIR spectra generally follow a linear relationship of intensity with respect to concentration, which is advantageous for determining new sample composition. Recent work has utilized multiple LR to identify heavy metals, including investigating the effect of surface chemistry on vanadium 47 and lead 48 toxicity. However, the simplicity of the method ultimately restricts the model's usefulness in more complex, dynamic systems.…”

Saccharide concentration prediction from proxy-sea surface microlayer samples analyzed via ATR-FTIR spectroscopy and quantitative machine learning

“…48 Absorbance FTIR spectra follow a linear relationship of intensity with respect to concentration, which is advantageous for determining new sample composition. Recent work has utilized multiple LR to identify heavy metals, including investigating the effect of surface chemistry on vanadium 49 and lead 50 toxicity. However, the simplicity of the method ultimately restricts the model usefulness in more complex, dynamic systems.…”