Calibrating predicted mixture toxic pressure to observed biodiversity loss in aquatic ecosystems

The supplementary files provided allow full reproducibility of the statistical workflows used to investigate the relationship between mixture toxic pressure (expressed as multi-substance Potentially Affected Fraction of species; msPAF) and biodiversity loss (expressed as Potentially Disappeared Fraction of species; PDF), including data input, analysis, and figure generation. Contents R Scripts: Each script corresponds to one or more figures or supplementary analyses in the manuscript. They are modular and organized to match the naming of input sheets in the associated Excel file (FIGURES DATA). Input Dataset (FIGURES DATA, Excel format): The Excel file includes multiple sheets labeled by figure number (e.g., Figure_2a, Figure_5, etc.) or analysis type (Statistical tests, msPAF_abiotic_factor analysis, etc.). These are directly referenced in the respective R scripts for reproducibility. The Info provides definitions for all variables. Output data (Oginah et al_SI2): The Excel file includes Table 1, which summarizes the ranges of modifying factors and mixture toxic pressure across different waterboards. It also includes Variance Inflation Factor results, statistical test outcomes, and predicted biodiversity loss (PDF) at regulatory and LCA benchmarks. The Info sheet provides definitions for all variables. The FIGURES DATA file Each sheet provides harmonized data to generate figures or conduct specific statistical tests. Data include: Chemical characterization data (e.g., use categories and frequency of detection across regions). Biota data (e.g., species-level abundance and site-level richness). Toxic pressure values (e.g., msPAF based on SSD-EC10, and for comparison, msPAF based on SSD-NOEC and SSD-EC50). Abiotic water quality parameters (e.g., pH, DOC, TSS, temperature), used in Variance Inflation Factor and sensitivity analyses. The structure of the Excel file ensures that: Each sheet can be directly input into its corresponding R script. Variable names are harmonized across sheets to allow merging and cross-referencing of datasets. Key analyses supported Calibration of msPAF-based toxic pressure predictions to observe biodiversity loss (PDF). Visualization of chemical use categories and site-level toxic pressure distributions. GAM modeling of species- and community-level responses to msPAF. Sensitivity analyses accounting for abiotic covariates using Variance Inflation Factors analysis (VIF). Comparison of results across regions (e.g., Netherlands vs. Delfland). Output data (Oginah et al_SI2): The info sheet contains the variable definitions Table 1 sheet contains the study output which includes: Waterboard identifiers and official names Water quality modifying factors (e.g., pH, TSS, DOC, temperature) range Mixture toxic pressure values ranges in the study area Results from the Variance Inflation Factor (VIF) evaluations for multicollinearity assessment Results from Kruskal–Wallis and Dunn’s post-hoc tests comparing species distributions on average across toxic pressure gradients Predicted and interpolated PDF values at regulatory (msPAF = 0.05) and Life Cycle Impact Assessment (msPAF = 0.2) benchmarks Summary outputs for national and individual water boards, including the difference in predicted biodiversity loss from baseline to benchmark levels