Nowadays techniques for non-targeted metabolite profiling allow for the generation of huge amounts of relevant data essential for the construction of dynamic metabolomic networks. Thus, metabolomics, besides transcriptomics or proteomics, provides a major tool for the characterization of postgenomic processes. In this work, we introduce comparative correlation analysis as a complementary approach to characterize the physiological states of various organs of diverse plant species with focus on specific participation of metabolites in different reaction networks. The correlations observed are induced by diminutive fluctuations in environmental conditions, which propagate through the system and induce specific patterns depending on the genomic background. In order to examine this hypothesis, numeric examples of such fluctuations are computed and compared with experimentally obtained metabolite data.