Abstract. Comparisons of the positions of species on Grimes'C-S-R triangular ordination model with their responses to individual environmental gradients indicates that the C-S-R model does not necessarily predict species ecological behaviour. The importance of the stress, productivity and disturbance gradients relative to other environmental gradients needs to be determined. In studies of species behaviour along a biomass/productivity gradient the collective vegetation property, biomass, has been confused with the environmental factor, fertility. Patterns of responses to biomass gradients e.g. Keddy's centrifugal model, should be examined in a two-dimensional environmental space to avoid such confounding effects. Assumptions regarding the shapes of species responses to environmental gradients remain untested. A recent model of species response functions to environmental gradients suggested that skewed responses curves show a pattern in the direction of the skew, always with the tail towards the presumed most mesic position on the gradient. Further evidence is presented to support this model for a temperature gradient in eucalypt forest in south-eastern Australia. 21 out of 24 species tested conform to the model.