Chapter24 Cichorieae Norbert Kilian, Birgit Gemeinholzer and Hans Walter Lack INTRODUCTION general lines seem suffi ciently clear so far, our knowledge is still insuffi cient regarding a good number of questions at Cichorieae (also known as Lactuceae Cass. (1819) but the generic rank as well as at the evolution of the tribe. name Cichorieae Lam. & DC. (1806) has priority; Reveal 1997) are the fi rst recognized and perhaps taxonomically best studied tribe of Compositae. Their predominantly HISTORICAL OVERVIEW Holarctic distribution made the members comparatively early known to science, and the uniform character com- Tournefort (1694) was the fi rst to recognize and describe bination of milky latex and homogamous capitula with Cichorieae as a taxonomic entity, forming the thirteenth 5-dentate, ligulate fl owers, makes the members easy to class of the plant kingdom and, remarkably, did not in- identify. Consequently, from the time of initial descrip- clude a single plant now considered outside the tribe. tion (Tournefort 1694) until today, there has been no dis- This refl ects the convenient recognition of the tribe on agreement about the overall circumscription of the tribe. the basis of its homogamous ligulate fl owers and latex. He Nevertheless, the tribe in this traditional circumscription called the fl ower “fl os semifl osculosus”, paid particular at- is paraphyletic as most recent molecular phylogenies have tention to the pappus and as a consequence distinguished revealed. Its circumscription therefore is, for the fi rst two groups, the fi rst to comprise plants with a pappus, the time, changed in the present treatment. second those without. The easy recognition of the members of the tribe Tournefort’s pupil, Vaillant, coined for his teacher’s comes along with a major drawback: the tribe is not only thirteenth class the name “Cichoracées” (Vaillant 1719) conspicuously poor in morphological features, but ex- and distinguished fi ve “sections” based on features of tensive parallel evolution of features further renders the the habit, pappus, and receptacle, the fi rst including all recognition of natural groups diffi cult. This situation has scapose taxa irrespective of their pappus, the second in- given rise to considerable diff erences in the generic and cluding those with a pappus of trichomes or scales and a suprageneric classifi cation of the members of the tribe by naked receptacle, the third those with a plumose pappus various students of Cichorieae. and a naked receptacle, the fourth those lacking a pap- Molecular phylogenetic studies have essentially im- pus, and the fi fth those with receptacular trichomes or proved our understanding of a few groups of the tribe paleae (Vaillant 1723; for an evaluation of Vaillant’s work since the 1990s. But only now the results of the molec- on Compositae see Greuter et al. 2005). ular phylogeny of a large dataset (428 taxa of 83 genera; Lamarck and De Candolle (1806) validated Vaillant’s Gemeinholzer et al., in prep.), representing the entire tribe, pre-Linnaean name for the tribe and subdivided Cich- have become available, and this has enabled us to provide orieae into four subtribes according to pappus features. The an essentially revised treatment of Cichorieae. While the lasting merit of these and the other 19th century authors 344 Kilian, Gemeinholzer and Lack dealing with the systematics of the Asteraceae in general Stebbins (1950), one of the key fi gures of the Modern and the tribe Cichorieae in particular, namely Cassini Evolutionary Synthesis, crowned his studies in Cichorieae (1827, 1830), Don (1828), Lessing (1832), De Candolle with a new subtribal classifi cation, based on a phenetic (1838), Bentham (1873), and Hoff mann (1890–1894), is multi-evidence approach by considering morphology (in their analysis, comparison, description, and classifi cation particular pappus, shape of the stigma branches, pollen, in species and genera of the enormously increased plant and indumentum), geographical distribution, and chro- diversity that successively became known to science in mosomal data (Stebbins 1953). In contrast to previous the course of this century, rather than their suprageneric classifi cations, Stebbins considered “each genus sepa- systems of subdividing the tribe. All attempts had in com- rately, placing it nearest to those genera which it most mon classifi cations based on one or a few key features, nearly resembles in respect to the largest number of char- pappus and receptacle characters having been particularly acteristics of external morphology, plus the nature of the highly appreciated (for further details see Stebbins 1953: chromosomes and the geographic distribution” (Stebbins 65–67). Extensive convergent evolution, especially in the 1953: 69). He arranged the 62 genera recognized by him pappus of the Cichorieae, however, condemned the re- into eight subtribes, thereby grouping genera with no sulting systems from Tournefort in 1694 up to Hoff mann pappus together with genera possessing a pappus, which, in 1894 to be largely artifi cial. however, resemble one another in other characteristics. Hoff mann’s (1890–1894) subdivision of the tribe, Within these groups the genera not always feature com- which had been infl uential until well into the 20th cen- mon characters but are sometimes united by transitional tury, illustrates the stagnation in the development of genera. Stebbins recognized the close affi nity of the en- the suprageneric classifi cation from the late 17th to the demic New World genera and placed them into two new late 19th century. He coined the name pair “Ligulifl orae subtribes, Malacothricinae and Stephanomerinae, which (Cichorioideae)” and “Tubulifl orae” (Hoff mann 1890– are distinguished by geographic distribution and chromo- 1894: 118) and separated Cichorieae as Ligulifl orae on some numbers. subfamily rank from all other tribes, which he united as Jeff rey (1966), in another phenetic approach under- Tubulifl orae. Hoff mann divided the tribe into fi ve sub- taken in the context of his studies of Cichorieae in tropi- tribes, of which his three larger subtribes are entirely cal East Africa, considered additional micro-morphologi- based on pappus features: Cichoriinae unite all genera cal characters (length of collecting trichomes on the style, without or with non-setaceous pappi, Leontodontinae trichome shapes on stigmatic surfaces, and pubescence of include all New and Old World genera with plumose the corolla tube), which he incorporated in his system to pappus, and Crepidinae include all genera with setaceous, improve Stebbins’s classifi cation. He defi ned groups and non-plumose pappus. In addition, he placed Scolymus in a subgroups but refrained from providing a formal taxo- subtribe of its own and united Dendroseris and Fitchia (the nomic classifi cation due to the “uncertain status of the latter actually an odd ligulifl orous Heliantheae; Carlquist ligulate Compositae within the family” (Jeff rey 1966: 1947) because of their arborescent life form. 428). Jeff rey’s classifi cation of fi ve groups, eight subgroups In the middle of the 20th century, a fruitful coopera- and eighteen series resulted in several natural groupings, tion of two American botanists, Stebbins and Babcock especially on the lower taxonomic levels. However, some- (Babcock and Stebbins 1938), revolutionized our under- times features are placed into a doubtful evolutionary con- standing of Cichorieae, as of plant systematics in general. text, e.g., he grouped the Scorzonera subgroup within the Studying the American species of Crepis, they discovered Hypochaeris group due to the paleaceous/plumose pappus the crucial role of hybridization and formation of poly- and medium to long style-arms, and the Crepis subgroups ploid complexes in the evolution of species. In the course within the Cichorium group due to long style-arms and of their subsequent cytological and taxonomic work in large collecting trichomes, not taking into account the Crepidinae s.l., they re-established and monographed possible diff erent evolutionary pathways by which these Cassini’s Asian genus Youngia (Babcock and Stebbins homologous characters could have evolved. 1937, 1943). Stebbins studied also the Asian Crepis rela- Bremer (1994) provided the fi rst cladistic analysis of tives Ixeris (Stebbins 1937c), Dubyaea and Soroseris, hereby the tribe, based on morphological characters, by studying making fundamental contributions to our knowledge of a selection of 23 from altogether 98 genera recognized, the vascularization of the ovary (Stebbins 1940). The pair which either represent presumed monophyletic groups, also investigated the genera Lactuca and Prenanthes (e.g., or distinct or isolated taxa. As result of this, he divided Stebbins 1937a, b) and provided a survey of karyology and the tribe in eleven subtribes, establishing the new sub- phylogeny in Cichorieae (Stebbins et al. 1953). By then tribes Catananchinae, Malacothricinae, and Sonchinae, Babcock had completed his monumental taxonomic revi- and left two genera, Cichorium and Scolymus, unassigned sion of Crepis, which takes karyological, morphological, to a subtribe. Due to the isolated position of Scolymus, and biogeographical data into account (Babcock 1947). he stated the necessity of a separate subtribe; however, in Chapter 24: Cichorieae 345 his treatment, monogeneric subtribes were avoided. For necessary), Lactuca (paraphyletic or polyphyletic depend- Cichorium he proposed a relationship close to Crepidinae, ing on circumscription,