Use of nonlinear regression techniques for describing concentration-response relationships of plant species exposed to contaminated site soils

Abstract The objectives of this study were to examine the effects of two contaminated site soils on seedling emergence and growth, compare the responses of different endpoints and species sensitivity, and develop appropriate statistical methods for the analysis of concentration-response curves. Plants were exposed to field-collected soils contaminated with amines or condensate. We reparameterized three nonlinear models (logistic, logistic with hormesis, and exponential) to determine any inhibiting concentration for a specified percent effect and confidence interval using regression analysis. Weighting procedures were applied, when necessary, to accommodate heteroscedasticity. This nonlinear regression approach was very satisfactory when used with data sets, each with 11 treatments, and produced an accurate, easily interpreted, and quantitative description of the data, which also provided qualitative information. The IC50s ranged from 2 to 96% contamination for condensate-contaminated soil and from 3 to 38% contamination for amine-contaminated soil. The responses were specific to species, endpoint, and soil. Mass measurements were generally more sensitive and precise than length measurements. Definitive tests were more sensitive than acute tests for endpoints other than emergence.