J. Hattori Bot. Lab. No. 76: 207-219 (Oct. 1994) SEXUAL REPRODUCTIVE CHARACTERS VS. MORPHOLOGICAL CHARACTERS IN LICHEN GENERA E. I. KA.RNEFELT1 AND ARNE THELL2 ABSTRACT. The generic concept and its early historical development until today's more natural multi-character based concept is discussed. The modern generic concept for lichenized ascomy­ cetes has traditionally been slightly different concerning crustose groups on the one hand and foliose and fruticose groups on the other. More studies of sexual reproductive st ructures in the large groups of lichens should presumably be carried out since it has been demonstrated recently that a considerable variation actually occurs in both asci and pycnidia in the cetrarioid genera. Some examples of difficult cases of evaluation of structural characters in relation to characters in the sexual reproductive structures in cetrarioid genera and the Teloschistaceae are also discussed. INTRODUCTION Almost two centuries ago during the classical period of botanical history, genera were basically described on simply recognized morphological characters or in a combination of the position of apothecia and pycnidia. Acharius ( 1803) in his Metho­ dus Lichenum separated 39 genera of which some are still valid, though in a different way, such as Alectoria, Cetraria, Evernia, Lecanora, Lecidea, Nephroma, Parmelia and Ramalina. Apothecial structures were also used later at the end of the classical period for the understanding of generic concepts. It was, however, not until in 1852 that spore characteristics were introduced by Massalongo for the recognition of genera. Massa­ longo even ranked characters in a certain order of importance, such as spores, asci, paraphyses, hypothecium, exciple and finally general thallus characters (Massalongo 1852). In all, Massalongo published more than a hundred lichen genera of which some are still recognized. His foremost opponent Nylander, however, ignored much of his work and instead emphasized the importance of anatomy and pycnidial characters. A large part of Nylander's impressive list of nearly 3000 newly described species and 83 genera is still valid. In some respects, the battle which was fought back in these days between Massalongo and Nylander and many others has continued to the present day, and no agreement has been reached regarding the most important characters for the separation of genera. MODERN GENERIC CONCEPT In modern times, morphological characters have continued to be used at generic level especially in fruticose and foliose groups, but generally a combination of anatom- 2 ' · Department of Systematic Botany, University of Lund, OstraYallgatan 20, 223 61 Lund, Sweden. 208 J. Hattori Bot. Lab. No. 76 I 9 9 4 ical and or chemical characters have been more important. In crustose groups, characters in asci, spore development and paraphyse structure have traditionally been considered of greater importance than morphology. We cannot point to any good modern example where a crustose genus has been recognized on morphological characters of the thallus only. However, numerous new genera which were formerly included in the Lecanoraceae and Lecideaceae s. !at. have been recognized not only on ascomatal characters but basically on diverging characters in the asci and hamathecium (Hafellner 1984, Hertel 1984, 1987). STRUCTURALLY-CHEMICALLY BASED GENERIC CONCEPT Chemical characters were already considered of major importance in the recogni­ tion of genera during the 1960s as in the case of the rather well-defined groups Cetrelia and Platismatia (Culberson & Culberson 1968). However, in other cetrarioid groups there is a relatively strong correlation between mainly morphologically based groups and secondary chemistry (Karnefelt & Thell 1993a, Karnefelt et al. 1992). In addition many recently recognized crustose genera such as Protoparmelia, Tephromela and Rhizop/aca were based on correlated chemical characters (Brodo 1986). Hale, who contributed considerably to the formation of a large number of genera in foliose groups was of the opinion that apothecial characters have far fewer or indeed no differences at all in larger groups. He was activelly searching therefore for new alternative characters which could support his systematic ideas. For instance, he emphasized the importance of thallus ornamentation such as the structure of rhizines and cilia (Hale 1974, Hale 1989, Hale & Fletcher 1990). Secondary chemistry of course became an important tool for supporting morphological characters (Elix 1993). Arcto­ parme/ia was recognized by a sparsely rhizinate under side in addition to the presence of alectoronic acid in the medulla (Hale 1986), Parme/ia was characterized by several combined medullary substances and effigurate pseudocyphellae (Hale 1987), Xanthom­ acu/ina by an umbilicate foliose thallus and strongly effigurate upper cortex (Hale 1985) andXanthoparmelia without atranorin in the upper cortex (Hale 1990). Many of these new or reevaluated names, which also include groups such as Pleurosticta characterized by the pored epicortex, cylindrical or sublageniform conidia and upper cortex HN03 negative (Lumbsch et al. 1988), however, have not been generally recognized for lack of strongly convincing correlated characters in the sexual reproduc­ tive structures (Eriksson & Hawks worth 1991). STRUCTURALLY-ANATOMICALLY BASED GENERIC CONCEPT In some genera, however, there is a clearer correlation between the general morphological characters and the anatomy of the cortex; for instance, genera such as Heppia and Peltu/a (Biidel 1987) and Psora and related genera (Timdal 1984) could partly be defined by their cortical structures. The genus Dege/ia was separated from Pannaria by the more periclinally arranged cortical hyphae (Arvidsson & Galloway 1981). Cortical structures have also been found to be of major importance in the definition of genera in the Physciaceae (Hale 1983). The larger forms of genera in the E. L. KARNEFELT & A. THELL: Sexual reproductive vs. morphological characters in lichen 209 1 2 3 ' Figs. 1- 3. Schematic drawings of asci. Fig. l. The Cetraria type, Cetraria islandica. Fig. 2. The Lecanora type, Melanelia form, Melanelia slygia. Fig. 3. The Teloschistes type, Xanthoria parietina. ab = axial body, rs = ring structure, oc = ocular chamber, th = tholus. Bar in Figs. 1- 3 = LOµm. Teloschistales, i.e. Caloplaca, Teloschistes and Xanthoria, can be defined on the com­ bined morphological anatomical character states, where the fruticose genus Teloschistes basically differs in its periclinally aranged cortical hyphae (Karnefelt 1989). Recently the genus Nimisia was mainly separated by unique characters in the cortical and medullary layers and well-separated from other possibly related groups (Karnefelt & Thell 1993b). SEXUAL CHARACTER BASED GENERIC CONCEPT During our project on the evolution and phylogeny of the alectorioid and cetrari­ oid lichens, the importance of sexual reproductive characters over mainly morpholog­ ical characters arose several times (Karnefelt & Thell 1992, Karnefelt et al. 1992, 1993). We arrived at the conclusion that ascomycete characters like ascus and hama­ thecial structures were of major importance in the evaluation of higher taxonomic categories on family and on generic levels. Furthermore we also included conidia in the same category of sexual reproductive characters, since these structures must mainly have a function as spermatia (Honegger 1984, Poelt 1986). We also concluded that major differences such as morphology and anatomy in general thallus organization cannot be neglected entirely since there must be strong genetical differences behind distrinctive structural changes in various groups. A large variation was discovered in the structure of the asci in the species which had traditionally been treated in the genus Cetraria s. str. Correlated mainly with characters in the pycnidia, in addition to anatomical characters in the cortex and secondary chemistry, the genus Cetraria could be better divided into several new and natural genera. The genus Cetraria was defined in the first place for 15 species with 210 J. Hattori Bot. Lab. No. 76 I 9 9 4 Figs. 4- 8. Asci in some cetrarioid lichens. Fig. 4. Asci of Arctocetraria nigricascens, characterized by a tholus (th) with a clearly broader axial body (ab) than in Cetraria, Russia, Jenisejsk 1876, Brenner 1827c (S). Fig. 5. Cetraria aculeata , ascus. Note the well developed apical ring structure. Sweden, Blekinge, 1871 , Swanlund (LD). Fig. 6. Ascus of Melanelia form with its typically, large axial body, here represented by Cetraria weberi, U.S.A., Arizona, 20. 05. 1987, Nash 22, 647 (LD). Fig. 7. Esslingeriana idahoensis, ascus of Tuckermannopsis form, U.S. A., Idaho, 1987, Anderegg 10 489 (LD). Fig. 8. Asci of Flavocetraria cucullata also with distinct ring structures (rs), Sweden, Oland, Andersson 1914 (LD). Bar in Figs. 4- 8 = lOµm. E. L. KXRNEFELT & A. THELL: Sexual reproductive vs. morphological characters in lichen 211 narrowly clavate asci with a ring structure in the tholus, oblong citriform conidia and a 2-layered cortical layer (Fig. S) (Karnefelt et al. 1993). Three morphologically similar groups were recently separated from Cetraria on anatomical and chemical grounds. The genus Arctocetraria, comprising two species, differs in a larger axial body in the tholus, paraphyses with very wide basal parts, bifusiform conidia and a 1-layered cortex (Figs. 4, 9). There are also chemical differences, since these species contain rangiformic and norrangiformic acid instead