ABSTRACT
Abstract .................................................................................................................444 30.1 Introduction ................................................................................................444 30.2 Materials and Methodology .......................................................................447 30.3 Results and Discussion ..............................................................................451 30.4 Conclusions ................................................................................................459 Acknowledgment ..................................................................................................460 Keywords ..............................................................................................................460 References .............................................................................................................460
The distribution of discrete ontogenetic states of individuals is usually spatially and temporally different within a population. If a sample from the population sample consists of several subsamples, the comparison of their ontogenetic spectra reveals heterogeneity of samples, i.e. different subsamples cannot be described by the same polynomial distribution. Therefore, the comparison of the samples using the aggregate data is not correct and tends to result in false inferences of biological importance. The paper proposes three methods for comparison of ontogenetic spectra of heterogeneous samples: a randomized variant of ANOVA, principal components analysis and ordinal regression analysis. The following approaches are exemplified in natural populations of cowberry Vaccinium vitis-idaea L. and epiphytic lichens Hypogymnia physodes (L.) Nyl. and Pseudevernia furfuracea (L.) Zopf.
