Multiple linear and principal component regressions for modelling ecotoxicity bioassay response

Area of Science:

• Environmental Science
• Ecotoxicology
• Aquatic Chemistry

Background:

• Aquatic ecosystem health relies on complex interactions between physical, chemical, and bacteriological factors.
• Understanding these multidimensional ecotoxicological interactions is crucial for effective environmental management.
• The Leça River's water quality and its impact on aquatic life require detailed investigation.

Purpose of the Study:

• To analyze the ecotoxicological response of Chlorella vulgaris and Vibrio fischeri to Leça River water quality.
• To identify key environmental variables influencing aquatic organism toxicity using statistical models.
• To provide a basis for evaluating the effectiveness of pollution abatement strategies in the Leça River.

Main Methods:

• Collected water samples from seven Leça River sites across five monitoring campaigns (2006).
• Analyzed 22 physicochemical and 3 microbiological parameters.
• Applied multiple linear regression (MLR) and principal component regression to ecotoxicity bioassay data.

Main Results:

• MLR model best explained ecotoxicity data.
• Chlorella vulgaris toxicity correlated negatively with dissolved organic carbon, zinc, manganese; positively with turbidity, arsenic.
• Vibrio fischeri toxicity correlated negatively with conductivity, turbidity; positively with phosphorus, hardness, iron, mercury, arsenic, faecal coliforms.

Conclusions:

• Identified specific water quality parameters significantly impacting aquatic organism toxicity in the Leça River.
• MLR provides a robust tool for understanding complex ecotoxicological relationships.
• Findings can inform targeted pollution control measures to improve Leça River water quality.