Morphological characters alone are extremely difficult to assess in Hieracium, especially in intermediate species and hybrids between basic and intermediate species. A morphological continuum of forms, further complicated by phenotypic plasticity, is difficult to deal with, and conclusions based on erroneously assigend material are evidently worthless. Furthermore, knowledge about origin, frequency and distribution of apomictic clones provides important information about speciation and population dynamics (Fehrer et al. 2002, Fehrer et al. 2003).
Classical DNA fingerprinting by Southern hybridization using mini- and microsatellite markers provide a useful tool for both purposes: preliminary morphological assignments can be regularly checked and some taxonomic problems could be resolved. Large differences in the frequency or origin and distribution between hybridogenous taxa were detected and related to different age and degree of fixation (Fehrer et al. 2005, Krahulec et al. 2004a, Krahulec et al. 2004b). The origin of invasive species could be inferred (Wilson et al. 2006).
Initial PCR-RFLP studies and sequencing of selected Pilosella species revealed that most genetic variability within the trnT-trnF region of chloroplast DNA is concentrated in the trnT-trnL intergenic spacer (Fitze & Fehrer 2000) on which we focused our further work. All Pilosella basic species fall into one of two major haplotype groups which are monophyletic and separated from each other by a number of homoplasy-free substitutions and indels. Within each group, several further indels (5-21 bp) are suitable for the distinction of subtypes by PCR-RFLP on high resolution agarose. These differences provide a useful tool for the distinction of the respective seed parents. Maternal inheritance of the chloroplast DNA in this group was ascertained by the analysis of progeny from reciprocal crosses using parents of known distinct haplotypes (Fehrer et al. 2005). Chloroplast haplotype distribution and patterns of molecular evolution across all Pilosella basic species suggest the differentiation of the two divergent haplotype groups in different glacial refugia (Fehrer et al. 2007b).
Within subgenus Hieracium, chloroplast DNA variability is generally low. Many different species share identical sequences. Haplotypes show no correlation to sectional assignment of taxa. Only in exceptional cases can presumed hybrids and hybridogenous taxa be distinguished by the above method. One of these examples is the recent identification of the seed parent in recent diploid natural hybrids between H. alpinum and H. transsilvanicum (Mráz et al. 2005). Also for this particular combination of taxa, maternal inheritance of cpDNA was proven by reciprocal artificial hybrids.
Sequencing of the trnT-trnL intergenic spacer of all basic species of subgenus Pilosella as well as of a number of Hieracium s.str. and Chionoracium species, of closely related Andryala, other Hieraciinae and some further Lactuceae taxa revealed - in combination with an ITS data set of the diploid species (B. Gemeinholzer) - that the current circumscription of the Hieraciinae is inadequate. In agreement with morphological characters (e.g. complex synapomorphic trichome types - diploma study of K. Krak), the molecular trees strongly support Hieracium s.l., Andryala, and the monotypic Hispidella hispanica as monophyletic and well distinguished from all other Lactuceae (Fehrer et al. 2003a, Fehrer et al. 2004). Discrepancies between chloroplast and nuclear DNA markers can be attributed to ancient wide hybridizations between the ancestors of different (sub)genera (Fehrer et al. 2007a). Ongoing research is aimed at a molecular phylogeny of the whole genus on a coarse taxonomical level.
Research