Reminder on data uploaded:
A Species Sensitivity Distribution (SSD) is a statistical approach used in ecotoxicology to estimate the relative sensitivity of different species to a chemical substance. The method compiles toxicity values such as LC50, EC50, EC10, or NOEC obtained from standardised ecotoxicological tests conducted on a range of taxonomically diverse species. These values are fitted to a statistical distribution, most commonly log-normal, log-logistic, or Burr type functions. From this fitted curve, a hazardous concentration for a given percentage of species can be derived. The HC5, which represents the concentration expected to affect 5% of the species, is widely used as a scientifically based threshold for ecological protection.
Regulatory authorities use SSDs to derive Predicted No Effect Concentrations (PNECs) and to establish environmental quality criteria for water, soil, and sediment. The approach is valued because it integrates interspecies variability in sensitivity and provides a transparent, reproducible method for risk assessment.
The OECD Test Guidelines (TGs) are internationally recognized and widely used methods for chemical hazard testing, encompassing tests across human health, environmental toxicity, and chemical properties. They are developed within the framework of the Inter-Organization Programme for the Sound Management of Chemicals (IOMC), involving organizations like WHO, UNEP, and OECD.
While many OECD TGs focus on endpoints like chronic toxicity or lethality tests (e.g., LD50 estimation), some guidance touches on modeling of quantitative binary or “quantal” data and dose–response analysis:
OECD Guidance Document on Statistical Analysis of Ecotoxicity Data (ENV/JM/MONO (2006)18) offers extensive coverage on statistical methods for ecotoxicity data—including quantal (binary) data modeling, dose–response modeling methods, hypothesis testing for NOEC (No Observed Effect Concentration), ECx estimations, and biological-based models.
Guidance Document 34 discusses methods like the LD50 up-and-down procedure (TG 425), focusing on estimating median lethal dose but does not characterize slope of the dose–response curve. Hence, it's limited in modeling binary dose–response slopes.
For broader testing contexts, Guidance Document 116 offers principles for dose selection in chronic toxicity or carcinogenicity studies (TG 451, 452, 453), valuable for designing tests—but not focused on binary outcome modeling specifically.
Recent OECD efforts include promoting Integrated Approaches to Testing and Assessment (IATA), including alternative methods like (Q)SAR, chemicals categories, and Adverse Outcome Pathways (AOPs), which may incorporate binary endpoints through predictive models, classification, or dose–response estimations.
In summary you can use this model for Survival tests: OECD 203 (fish), 202 (daphnia immobility), 210–212 (fish life stage), 207 (earthworm), 401/420/423/425 (rodent oral), 223/205 (birds). Mobility tests: OECD 202 (Daphnia immobilisation), some soil invertebrate avoidance assays. Cell survival tests: OECD 129, 249, 442 series, 455 (in vitro cytotoxicity and mechanistic endpoints).
OECD Guidance Document - ENV/JM/MONO (2006)18 Statistical analysis of ecotoxicity: methods for quantal (binary) data modeling, NOEC, ECx.
OECD GD-34 (ENV/JM/MONO (2005)14). Up-and-down LD50 procedure; binary outcome but slope not modeled
OECD (2014). Guidance Document 116 on the Conduct and Design of Chronic Toxicity and Carcinogenicity Studies, Supporting Test Guidelines 451, 452 and 453: Second edition OECD Series on Testing and Assessment, No. 116, OECD Publishing, Paris.