The normal distribution as appropriate model of developmental instability in Opuntia cacti flowers.

Developmental instability (DI) as measured by fluctuating asymmetry (FA) has been proposed to reflect fitness and stress. Furthermore, the associated developmental buffering may reduce morphological variation, conceal the expression of genetic variation and as such play an important role in evolutionary biology. However, observed associations between FA and various forms of stress and quality appear very heterogeneous. Presently it is difficult to interpret the biological relevance of this heterogeneity because little is known about the link between FA and the underlying process of DI, casting doubt whether DI can be viewed as an individual property and how closely FA reflects the underlying process of DI. Therefore, studies that explicitly test the validity of assumptions of the proposed theoretical models and estimate between-individual variations in DI are needed. We present data on Opuntia cacti floral traits confirming that the normal distribution can be viewed as an appropriate approximation of the distribution of DI and that the concept of hypothetical repeatability can provide useful insights into the interpretation of patterns in FA as a measure of DI. Furthermore, we detected significant between-individual variation in DI. Measuring petals from several flowers within individual plants allowed making inference of individual DI.