Establishing ecologically-relevant nutrient thresholds: A tool-kit with guidance on its use

Area of Science:

• Environmental Science
• Ecology
• Statistical Modeling

Background:

• Eutrophication management requires realistic thresholds to protect ecosystem services.
• Nutrient impacts on biota are often influenced by multiple co-occurring stressors.

Purpose of the Study:

• To introduce a toolkit of statistical approaches for setting nutrient thresholds.
• To explain the rationale and applicability of different statistical methods under various conditions.

Main Methods:

• Regression analysis for strong nutrient-biota relationships.
• Categorical approaches (binomial regression, classification mismatch) for weaker relationships.
• Multivariate approaches (decision trees, quantile regression) to account for multiple stressors.

Main Results:

• Regression is effective when nutrient-biota relationships are strong.
• Categorical methods are suitable for weaker relationships but may underestimate thresholds with co-stressors.
• Quantile regression can establish thresholds irrespective of other pressures.

Conclusions:

• The toolkit provides useful statistical methods for eutrophication management.
• Understanding ecological processes and local context is vital for threshold setting.
• Sophisticated, water-body-specific approaches may be needed alongside general methods.