Introduction: Molecular systematics in lichenology - quo vadis ?

The introduction of molecular methods in lichenology was accompanied by hopes that sequence data would be the answer to problems of traditional lichen classification. After nearly a decade there is a wide range of such molecular approaches. An on-going debate is focused on the origins of lichenization within the Eumycota. Is this derived from a single gain of lichen-habit, which was lost numerous times in more advanced groups of fungi, or from multiple lichenisations ? Results from different working groups are contradictory, which suggests that more data are required. Nucleic acid sequence data are also used in studies at lower taxonomic levels. Current discussions on generic concepts will be stimulated by molecular data, and this also applies to "species pairs" and the relationships of taxa with uncertain systematic position. The advent of PCR facilitated molecular work with lichens, and sequencing will become a standard technique in lichenology in the near future as 'wet' laboratory techniques are now fairly optimized for work with lichens. DNA-isolation is possible from a very little amount of material and the availability of non-algal primers allows amplification of fungal target DNA. It is not a problem to work with herbarium specimens of certain groups. However, species exist where sequence data are obtained only with difficulty, and prior to DNA-isolation, careful examination of the samples is essential to avoid undesired amplification of contaminating fungal material. The first gene studied with a broader range of lichenized fungi was the nSSU rDNA, which codes for the smaller nuclear subunit ribosomal RNA, an indispensible part of the protein synthesis apparatus. SSU rDNA has been studied extensively by molecular biologists who have accumulated a huge amount of information about the function of this molecule. Particularly in lichen mycobionts, numerous insertions (many of them are group I introns) have been found, and knowledge is growing about the position and nature of the insertions which are excluded from phylogenetic analyses of the data. Ribosomal RNA is a structural molecule with a complex pattern of secondary and tertiary intramolecular interactions, including helical regions and pseudoknots. The constraints on a particular higher order structure may result in frequent complementary changes of paired nucleotides. Although such characters are not independent, informative characters for phylogenetic analyses are found by comparing an alignment of one-dimensional sequence data. SSU and LSU rDNA are mainly used to study relationships at higher taxonomic levels as these genes are quite conserved in general and only contain short regions of high sequence variability. Within the ribosomal gene cluster there are also sequence regions with a low degree of conservation. The ITS (Internal Transcribed Spacer) regions between SSU and LSU rDNA are a good example. In mycology, the ITS genes have been used in studies at low taxonomic level, for example investigations of species relationships. However, they may also be of interest at higher taxonomic levels. For higher plants it has been shown that the ITS2 region can be used to resolve the relationships of angiosperms (Hershkovitz & Zimmer, NAR 24, 1996). Sufficient sampling of taxa is necessary in such approaches and even then, alignments of such highly variable sequences may be problematic. A future alternative to ribosomal genes will be protein genes. Since these follow a 3-bases reading frame, less ambiguity in the alignments is expected even if distantly related taxa are compared. There is no consensus about the most appropriate method for sequence data analysis. Phylogenetic analysis is usually carried out using parsimony. However, Gargas (Am. J. Bot. 84, Suppl., 1997) showed that maximum likelihood was a more robust analysis than parsimony and slightly different topologies result from analyses of the same data set with either parsimony or maximum likelihood. Maximum likelihood is a computer-intensive technique and handling larger amounts of data will be problematic without use of high-performance hardware. A different approach to classification by using signature sequences within the rDNA data was proposed by Eriksson (Can. J. Bot. 73, Suppl. 1, 1995) to circumscribe larger taxonomic units. Signature sequences are short pieces of sequence characteristic of a taxonomic group. This approach is probably the most traditional in molecular systematics, because it principally adds molecular traits to traditional classification without phylogenetic analysis. Most molecular investigations are carried out with single genes of the rDNA cluster, therefore phylogenetic trees only represent the phylogeny of genes. This is done by assuming a constant mutation rate in the genes and interpreting sequence variation as a kind of chronicle of evolution. Today, the only way to discuss both the validity of these assumptions and the results of molecular studies is to compare them with traditional concepts. It is frequently observed in phylogenetic studies of rDNA that certain taxa do not have a stable or well supported position in a tree. This is often due to a comparatively high sequence divergence in those taxa and may indicate insufficient sampling of taxa or molecular characters. Two strategies may help in this situation: gathering more taxa and more sequence information per taxon. More sequence information will include combined analyses of sequence data from more than one gene. Lutzoni et al. (Am. J. Bot. 84, 1997) showed under what conditions different molecular data sets may be used in combined analyses. Issues of taxon sampling have been clearly pointed out by DePriest et al. (Am J. Bot. 84, 1997), who suggest that a group in an analysis should be represented by at least 4 taxa. Phylogenetic trees are strongly dependent on the set of taxa which is included in an analysis - both as ingroup or outgroup assemblage. Also important in the interpretation are the gaps between the clades, i.e. the taxa which have been excluded for certain reasons but which may have considerable influence on the tree topology. Furthermore, one may question whether sampling of taxa should be more detailed in presumably older groups with longer evolutionary history. Finally, if we feel certain of a phylogenetic hypothesis, how can this be transformed into a classification? It has been suggested that classification systems without ranks can be applied in fungal systematics. My feeling is that this will cause practical problems, at least with the current state of knowledge. Given the fact that innumerable Ascomycetes are still to be described or revised in very traditional but urgently needed monographs, it is premature to adopt a rank-free classification. Despite these considerations, the next years will probably be the most interesting for molecular lichenology as the field emerges from infancy and we learn from our experiences. Real progress will be achieved by the symbiosis of traditional and molecular approaches.

Lichenology will be improved by molecular systematics!

Change is difficult, but none the less beneficial. As the twentieth century ends, there is a sense that lichenology as a field hesitates to move forward. It is all too easy to erect barriers and blockades to progress, perhaps an uneasy consequence of our collective obsessions. We agonize over field-specific issues such as the recognition of chemotypes as species, and the segregation of genera until they contain 20 or fewer species. Should we recognize species pairs? Can ascal tips provide the key to familial classification? Hypotheses have been advanced, data have been gathered, but definitive conclusions remain elusive. Instead of focusing on recognizing species, genera and families, we ask again and again what arbitrary rules allow us to find them? Have we forgotten that lichen species OUTNUMBER sexual species of Ascomycetes? Concerning fungal classification, lichenologists can seize the upper hand. Will they? Many of the character systems developed for lichen-forming fungi have no equivalents in other fungal groups, making lichenology a quaint and eccentric relative to ‘normal’ mycology - an amusing uncle no one takes seriously. This eccentricity, though not without its charms, constrains lichenology as the poor relation of mycology, and even most of the traditional botanical disciplines, in funding and staffing but especially in status and standing. With molecular tools lichenology has the potential not just to catch up, but to be at the forefront of research among these and other disciplines. It is time to cast off the ‘poverty-mentality’ of our past - DNA is a great equalizer of large and small organisms, of the culturable and the obligately symbiotic. This new era may be said to have begun five years ago at the IAL2 in Lund, when a half-day symposium marked the beginnings of the molecular subdiscipline in lichenology. Today, molecular research is no longer a privilege but is an expectation for young scientists, even in lichenology. Talks on molecular results could go on for days (or at least seemingly so to those who are a bit wearied by the specialized terminology). With refined molecular tools lichens have proven their broad potential to answer questions central to the whole of biology. Lichenology has already made a direct attack on fundamental questions, not least those with evolutionary and phylogenetic components. With this lichenology has led the systematics community in modernity and sophistication. Now is the time to state clearly: what important results are gained through molecular lichenology? How has lichenology realized the promise of molecular techniques? It is time to stop calibrating molecular techniques against traditional classification and systematics and use our results to present new and exciting hypotheses. Or will we show ourselves as a rebellious new generation using current fungal classifications as our strawmen and traditional taxonomists as our whipping boys? Molecular systematics research done by lichenologists, and including lichen species, has already shown that the lichen habit evolved multiple times and in diverse groups of true fungi. Phylogenies based on molecular data, including lichen-forming species of fungi, have resolved relationships of the Ascomycetes, and shown that the old classes Discomycetes, Plectomycetes and Pyrenomycetes are not monophyletic, and should be abandoned. Furthermore, molecular data have clearly demonstrated that lichen species are highly variable. Some may find these results trivial (see Jørgensen, Intern. Lichenol. Newsl. 28, 1995), yet they have stimulated interest among mycologists, botanists, and biologists in general. Results based on molecular data have gained attention for lichens and lichenology in the international science world and the popular media. One arresting new observation, in the discovery of which we were privileged to take part, is that lichens (at least those in the order Lecanorales) have a remarkable incidence of group I introns in their ribosomal genes. This finding has aroused interest from evolutionary biologists, medical researchers and even those in theoretical biochemistry. What do these group I introns mean, and how can lichenologists run ahead with this new research area? Our advances in lichen-based research have not gone unnoticed by the people with the authority to enable further lichenological research through awarding jobs and funding. Most importantly these new techniques have allowed us to showcase lichenology and lichen-forming fungi as lucrative research areas. Perhaps a decade ago a professor in mycology would have been reluctant to take on a student interested in lichen research. Today mainstream "mycologists" actively encourage projects on lichens. There is a particular satisfaction to being at the frontier of a field - teetering near the edge. Although as a reasonable biologist, one will have moments of doubt. All leaders are vulnerable to attack. Do molecular advances have to be paid for with competition and conflict? We are sadly afraid so. Lichenology is a very small pie in the overall scheme of things, and lichenologists presume that there is not enough fame and fortune to go around. Instead of fighting over who gets the bigger piece of the pie, it would be more profitable to seize a bigger pie, so to speak, by expanding lichenology and lichenological topics. Unfortunately it is easier and more scientifically defensible to deconstruct vigorously competing hypotheses than to welcome serendipity and truly novel discoveries. There is admittedly some conflict among current phylogenetic results, and questions remain unanswered. Which results most closely approximate reality? In this we know that only hindsight will provide us with clear vision, and before we know what is ‘true’ we are obliged to forge ahead. Among other phylogenetic results there is common agreement, and this should be emphasized for the short-term. We agree that Ascomycetes are a monophyletic group, we agree that the Pezizales are basal among the filamentous Ascomycetes (excluding the enigmatic genus Neolecta). For that which remains in question, seemingly wasteful gathering of the same sequences and repetition of the same analyses for independent confirmation is a necessity of scientific progress. Only through such painful and uncertain exercises can we ever hope to reach true consensus. Molecular lichenologists are now poised to publish their results as DNA sequences and phylogenetic trees even if these hypotheses are preliminary. Proposed classification schemes will serve as strawmen for the next generation of research. These molecular classifications should be rigorously tested against new molecular data and phylogenetic analyses and, what is most important, against anatomical or morphological character systems. At the IAL3 in 1996, our research group first presented a molecular phylogeny based on small subunit ribosomal DNA (SSU rDNA) for the suborders of Lecanorales in poster form. In two manuscripts, one currently submitted and one in preparation, we will propose that the Lecanorales includes the suborders Agyriineae, Cladoniineae, Lecanorineae, Peltigerineae, Teloschistineae, and the family Sphaerophoraceae, but excludes the suborders Umbilicarineae, Pertusariineae, possibly Acarosporineae, and the asexual taxa Siphula and Thamnolia. Will this phylogeny be wrong? Absolutely! Will the ribosomal DNA, and any other gene we could use, have its own evolutionary noise? Certainly! But these phylogenies will serve well as a strawman to stimulate future work, just as the multiple origins of the lichen habit already have! It seems that as a field the molecular phylogenetics of lichen-forming fungi has entered its difficult teen years. Gone are the blushful days of naïveté; we know that molecular data will not solve all problems instantly. A few tearful tantrums are to be expected during the maturation of molecular lichenology. Molecular systematics appears deceptively simple, but data analysis and interpretation remain fraught with complications. Data accumulate faster than they can be processed, and some may have to be reserved to await new methodological and technological tools. We are in the enumeration phase. As with any innovation in systematics, cytology, chemotaxonomy or phytogeography, an initial phase of promise in these subdisciplines was necessarily followed by an era of datacataloging and management before true progress could be registered - much as Chicita Culberson’s "Chemical and Botanical Guide to Lichen Products" (University of North Carolina Press, Chapel Hill, 1969) was critical in developing lichen chemotaxonomy. Each sequence submission to repositories such as Genbank and each published phylogenetic hypothesis adds to our comprehensive reference library. In truth, lichenology and lichenologists cannot turn back from the molecular revolution - that would signal oblivion for our field. We are obligated as scientists to move forward with alacrity and vision tempered with a realization of our limitations. This, in combination with a sense of wonder, a feeling for our organisms, and a healthy dose of humor will stand us in good stead to meet the challenges of the future.

Reactions

Both contributions were very stimulating and instructive, even for a layman in this field as I am. I did not, however, like very much the spirit of the introduction to De Priest's contribution: is it really necessary for "molecular" people to show the world that their data are THE data, and that everything else is conservative, outdated and not worthy of serious consideration by "progressive" scientists? We have heard this so many times in the past, whenever a new technique has appeared on the market! In Biology the study of relationships is essential, and taxonomy, in particular, has always been and always will be the converging point for data coming from widely different disciplines. Some of the "field-specific" issues mentioned by De Priest are, on the contrary, fundamental questions for Biology as a whole. The sentence "...instead of focusing on recognizing species, genera and families, we ask again and again what arbitrary rules allow us to find them" is something I cannot understand: how can we "find" taxa if we do not agree on their operational definitions? Molecular data are certainly important for solving biological problems, but the idea that they should be granted a kind of special status is, in my opinion, far from being a "progressive" one. On the contrary, it is a narrow-sighted attitude which in Biology has already produced true disasters by bringing fundamental disciplines to the verge of extinction in some countries. Maybe this attitude reflects the pathologically harsh concurrence for funding which characterizes the American system: mors tua, vita mea. I do not think that this kind of Social Darwinism is healthy for a sound development of such a complex discipline like Biology. We should work together, trying to understand each others' problems, and to relate each others' results: finding a relationship between molecular data and other data coming from classical ecology, morphology, physiology is the best possible investment we can imagine for our results: in this way our data gain an enormous "added value". To do this, we will always need ecologists, morphologists, physiologists etc. Life is much more complex than the structure of nucleic acids !

All characters which are of interest for taxonomists, from morphology to ecophysiology, ultimately derive from the structure of nucleic acids. Contrary to Le Bois, I can well conceive a taxonomic system based exclusively on molecular data. When will this happen ? This is another story...

There are about 13500 species of lichens described using morphology according to the "Dictionary of Fungi", and they serve as the ultimate basis for any molecular approach. The turnover of names suggests that there is still a lot of traditional work needed to improve molecular work. A funding policy which gives a lower priority to non-molecular lichen systematics is highly questionable. It might end up in molecular phylogenies of the genus Lichen! Most molecular studies are presented at congresses as being in preparation or in press. During the last 3 years everybody realized that hypotheses from different working groups who are studying the origins of lichenization are highly divergent. If sequence data are THE data, what is the reason for this incongruity? On what results can we rely? Shall we not trust anything at all or will we have to accept "multiple origins of taxonomies" as suggested by US Science? It might be interesting to merge the data sets of all the competing phylogenies. In molecular systematics all hypotheses are developed using a DNA-sequence alignment of a selected set of species (selected by what criteria?). Based on this alignment, phylogenetic trees are produced by applying a defined model of evolution. If we agree on the reproducibility of an alignment, will we still get highly divergent hypotheses, even when the sets of sequenced organisms overlap? Do we have to ask whether the evolutionary model is sufficiently well adapted for analyses of ribosomal genes? How to test this? If questionable results are produced, it will also be important to review the material used for sequencing. Storage of used material in herbaria, and an indication from where mycelium has been taken from the specimen must ensure this. Many lichenicolous fungi in certain lichen groups may lead to odd sequences. Indeed, "progressive" scientists may improve "traditional" knowledge of lichenicolous fungi by molecular approaches to the host lichens.

First of all, I thank all contributors for this timely debate. I feel quite optimistic about the future of molecular biology in connection with lichen biology and phylogeny. This is not just a technique to confirm phylogenetic trees based on morphological or other kinds of characters. It is a novel approach which, surely, has to incorporate other information but will, in the long run, provide an unifying view of previous understanding. Two subjects, at least, deserve being emphasized: lichen individuals and lichen populations. The study of lichen individuals is a complex subject that will certainly benefit from an understanding of the molecular biology of the single thallus. Furthermore, the molecular study of lichen populations is not an easy subject, and a trial and error approach will be mandatory to find reliable markers of populations. In my view, there are excellent prospects of incorporating the methods of molecular biology into the current effort of using lichens as bioindicators of atmospheric pollution, particularly in the study of the re-colonisation taking place in some parts of Europe following the improvement of air quality. Some exciting problems, such as the relationship between genotype and ecophysiological characteristics of populations, or that between biogeographic and genetic variability, can greatly benefit from the incorporation of molecular data. The most important contemporary paradigm cannot be but utterly rewarding for the development of our peculiar kind of mycology.

The PCR or ADN-ADN hybridization techniques result in genetic distance. Are such data useful for reconstructing phylogenies ?

Topic 2: CITATION OF SPECIMENS AND OF WEB PAGES

There is a major problem that has been developing in monographic studies. The editors of journals will not let us cite many specimens seen. To me, the list of cited specimens is over half of the value of a monograph! I can look at the cited specimens and check the herbarium for duplicates, so I know what the monographer means. Now, I can put data bases or list of all specimens seen on the Web, but how does anyone ever cite a Web page? Is it a valid citation? If I look at someone's Web page and see that they have recorded these X species from X park, can I cite the page or do I just have to ignore that valuable information? The citations can be made available on the Web but why do it if none can use it and cite it? Any suggestions?

Topic 3: PLACYNTHIELLA OR SACCOMORPHA? THIS IS THE PROBLEM...

What is the right name for "Lecidea uliginosa group" - Placynthiella Elenkin, Placynthiella Gyelnik or Saccomorpha Elenkin ?

The Saccomorpha-Placynthiella story is perhaps worthy of a brief discussion among us. Nimis & Poelt (1987, Lichens from Sardinia: 218-219) translated the original "description" of Placynthiella from Russian (Elenkin 1909). Most of the text is a description of sand dunes. Some "black crusts" are mentioned, attributed to "Placynthiella arenicola Elenkin nov.sp. et nov. gen.", with a note saying that this lichen will be treated in a forthcoming article. A few years later, Elenkin (1912) formally described this lichen under the name "Saccomorpha", with a very detailed description, and adding that it was already "mentioned" by him under the name Placynthiella in his earlier paper - thus showing that he did not consider Placynthiella as a validly described taxon. According to Nimis & Poelt (1987) Placynthiella is a nomen nudum. Presently, we have the unfortunate situation where north European authors still use Placynthiella, while south European authors use Saccomorpha. In one way or another we should agree on a name, whatever that will be. The use of Placynthiella implies the recognition of the sentence "black crusts" as a valid description; I am not an expert nomenclaturist, and it could be that this is OK (here advice is needed by the few colleagues who are real experts in nomenclatural codicils). However, my very personal feeling, based on common sense, is that if the "description" of 1909 is a valid one - and this against the real intention of the author - then I am the Emperor Josef II of Austria!

We agree with Nimis & Poelt (1987) that Saccomorpha Elenkin is the correct genus name, given that the earlier name Placynthiella Elenkin was a nomen nudum (inadequate description). Coppins, James & Hawksworth (1987) considered Placynthiella Gyelnik a synonym of Placynthiella Elenkin. Unfortunately they are incorrect in accepting Placynthiella Elenkin (1909) as validly published, and the name Saccomorpha (1912) takes priority. The only remaining question is the identity of Elenkin's type species S. arenicola - Hafellner (1984) considered S. arenicola a synonym of S. uliginosa (Schrad.) Hafellner, but more recent examination of the type by Coppins, James & Hawksworth equates P. arenicola Elenkin with P. hyporrhoda (Th. Fr.) Coppins & P.James. - Saccomorpha Elenkin, Ber. Biol. Süsswasserstat. Kaiserl. Naturf. Ges. St.-Petersburg 3: 194, 1912 (Reprint at US!). Syn. Placynthiella Elenkin, Bull. du Jardin Imper. Bot. St.-Petersburg 9 (1): 18, 21. 1909, nom. nud., Placynthiella Gyelnik, Annls. Hist.-nat. Mus. Natn. Hung. 32: 187. 1939. Type species, S. arenicola Elenkin. [=S. hyporhoda (Th. Fr.) Clauz. et Roux, fide Coppins, James & Hawksworth (1987)].