MOSAIC_SSD: a new web tool for species sensitivity distribution to include censored data by maximum likelihood

Area of Science:

• Environmental toxicology
• Ecotoxicology
• Ecological risk assessment

Background:

• Censored data are prevalent in species sensitivity distribution (SSD) analyses but often excluded, leading to reduced dataset size and representativeness.
• Excluding censored data can result in less reliable ecological risk assessments and predictions of hazardous concentrations.

Purpose of the Study:

• To develop and present a method for incorporating censored data into SSD analyses.
• To introduce the MOSAIC_SSD web tool for fitting SSDs with censored and uncensored data.
• To demonstrate the added value of including censored data for improved hazardous concentration predictions.

Main Methods:

• Utilized an extension of the standard maximum likelihood method to accommodate censored data in SSD.
• Employed the R-package fitdistrplus for fitting parametric probability distributions to ecological data.
• Developed the MOSAIC_SSD web tool for comprehensive SSD analysis, including censored data handling.

Main Results:

• The MOSAIC_SSD tool successfully fits SSDs to datasets containing both censored and uncensored data.
• The tool provides predictions of hazardous concentrations with associated bootstrap confidence intervals.
• Including censored data in SSD analysis demonstrably enhances the representativeness and accuracy of predictions.

Conclusions:

• Integrating censored data into SSD analysis is feasible and significantly improves risk assessment outcomes.
• The MOSAIC_SSD web tool offers a practical solution for utilizing all available data in SSD.
• This approach leads to more robust and reliable predictions of environmental hazard concentrations.