Multivariate quantitative structure-activity relationships for the aquatictoxicity of alkyl polyglucosides

The aquatic toxicity of 34 alkyl polyglucosides (APGs) towards two fresh-wa ter species, Thamnocephalus platyurus and Brachionus calyciflorus were stud ied. The toxicity tests were performed using so-called toxkits, and for eac h surfactant the results are presented as (10)log (mean LC50) values. The t oxicity data were combined with physico-chemical data for the APGs, and a M ultivariate Quantitative Structure-Activity Relationship (M-QSAR) model was calculated. Partial Least Squares (PLS) regression was used to develop the M-QSAR model. The resulting linear M-QSAR model explained 93.6% of the var iance in the biological response and had a predictability of 86.6% accordin g to cross-validation. The physico-chemical properties with the strongest i nfluences on the toxicity of the surfactants were the critical micelle conc entration (c.m.c.), wetting, contact angle, and number of carbon atoms in t heir hydrophobic parts (C and redC).