Topic: Genus concepts in lichenology, proposed
by P.L. Nimis
The present tendencies concerning the genus concept in lichenology somehow recall the Italian-Slesian revolution of 150 years ago, based on the use of microscopical characters. At that time, however, it did not end well: the hasty creation of hundreds of names was threatening nomenclatural chaos, and this just when new lichens brought back from the colonies awaited to be rapidly filed into clear-cut generic containers by the few lichenologists of the newly created National Museums. The harsh reaction started by Nylander culminated with the pinnacle of Zahlbruckner's Catalogus, where extremely unnatural generic concepts were often adopted. This was indeed a real monument, but also a heavy sarcophagus for fresh, sound, evolutionary ideas. The sarcophagus was so imposing that it conspired to hold back the recognition of monophyletic units amongst lichen-forming fungi for over a century. Timid approaches to resurrect a few more natural genera appeared only a few decades ago, first mainly at the expense of some of the larger crustose mega-genera, then also of some macrolichen genera. It was like a small rivulet of new genera, growing to a larger and somehow solemn river: Lichenology was waking up from an all too long sleep. Then, suddenly, came the flood, announced by a first wave: the splitting of Parmelia. No doubt, these studies brought about an immense increase in the knowledge of this group. However, several lichenologists felt, and still feel, that most of the new generic segregations were not justified, and that once this trend received more widespread acceptance, it would have resulted in serious problems for lichen taxonomy. This as an introduction. Here are some points for discussion: 1) Some recent generic splittings are often justified on practical grounds: the old genera were "too large". My objection is that, when both large and small genera do exist, this fact should be accepted and recognized nomenclaturally ("If God created large genera, why should men go against God's will?"). Otherwise, examples of "practical" splitting of large genera will be automatically extended to small genera as well, as is happening today, and this is much more difficult to justify on "practical" grounds. 2) "God created only the species, and not the genera", somebody objected during a dinner in Salzburg. I maintain that the Creation has an inner structure, which means that God created also the genera (please, have a look to any cladogram if you do not believe this). Hence, we cannot do just what we want with generic concepts, and if we split large genera just because they are large...the Inferno is waiting for us. 3) More and more lichenologists tend to assume that every supposedly monophyletic group of species deserves to be treated at the rank of genus. As there is no "lower limit" to the definition of the term "monophyletic" this assumption could lead to the recognition of every single species as a genus, which is obviously an absurdity. Where should we stop splitting? 4) Taxonomy has two faces: one for specialists, the other for users - and users are many more than specialists can see from their ivory towers. Generic names are an integral part of the binomial, and any generic change has heavy repercussions on nomenclatural stability. This does not apply to subgeneric-supraspecific ranks. When the taxonomy of a given group is far from being settled, and if there is no clear evidence that a supposedly monophyletic group of species has nothing to do with its older genus, the tentative segregation should be done at subgeneric rank. The new taxonomic information will be there, without causing unnecessary, and often provisional, nomenclatural changes. Names are too important and delicate a matter: we cannot play with names. 5) Do we agree that lichens are fungi? More attention should be paid to generic concepts outside the narrow field of lichenology. Current trends risk isolating lichenology from Mycology again.
I myself had the following nightmare last year: "Let's take a large lichen family, with two small macrolichen genera and a huge genus of crustose lichens (Teloschistaceae). I can start from the small genera, segregating one species here, two species there, because of small anatomical differences. Why not? It has been already done elsewhere. Now the problem arises: what will this mean for Caloplaca? Well, some more or less clearly-defined groups are there, attracting, like mermaids, suitable candidates for new genera. No doubt, the new segregations could work, at least for a while, because a huge container is left for all the rest. Let's do it...". After waking up, I realized that, before doing it, one should first explore the whole container, since several surprising things could be left there. If I hastily try to slice from Caloplaca some juicy beef-steak genera, I'll be likely to create the premises for a nice hamburger. Then we'll have to deal with a nomenclatural hamburger. Gaining what in terms of new knowledge? Not a jot. Fast-food taxonomy? No thanks.
Many lichen genera are still artificial, and deserve careful segregation. Within larger units, such as Caloplaca, or the Verrucariaceae, we still have to tentatively define smaller groups of closely related species: is it really necessary to crown the recognition of every small supposedly monophyletic segregate with a new generic name? Should the motto of modern lichen taxonomy be: "If I don't do it, somebody else will do it"? Should not what happened one hundred years ago be a lesson for all of us? The whole matter is complex and controversial, it urgently needs discussion, and nobody can claim to have the right answer. In Salzburg several new generic segregations smoothly melted away before a more or less silent audience. In this, at least, history does not repeat itself: how far was that all too quiet pond from the stormy discussions about generic concepts of 150 years ago! But this, I am afraid, is not exactly a sign that the old sentence is true: "Historia magistra vitae".
Pier Luigi Nimis, Trieste
Reactions
The whole discussion on a "common generic concept" is stimulating but somewhat futile because there are no means to agree upon such a concept. First of all, genera or higher taxa do not evolve. They are human abstractions and not units of evolution. Only species evolve, or some other unit such as populations that can be conceived as existing entities of nature. Secondly, phylogenetic systematics is science, and any practical purposes of systematics such as nomenclature are subordinate to science. To be scientific, observations must be repeatable and hypotheses must be falsifiable. In working with phylogenetic systematics there are some basic assumptions that we must accept: 1. Organisms form patterns of groups that can be hierarchically arranged. 2. This hierarchy can be discovered by grouping the organisms on shared, derived characters through character state changes. 3. Ad hoc hypotheses of homoplasy are minimized by using the parsimony criterion. The diagrams, the phylogenetic trees, produced under those assumptions represent the hierarchy. The branching points, if being used as taxonomic groups, must include all members of the group otherwise groups will not be monophyletic. Therefore a suggested classification must be in accordance with the phylogeny. That includes the monophyletic groups we choose to call genera. There is no way round this. However, the size of a genus, how many species it should include, is a (subjective) decision that the individual taxonomist has to take.
Anders Tehler, Stockholm
I agree with Pier Luigi's introduction to the Forum discussion. Two main points should be addressed: 1. The grouping of taxa, and, 2. The ranking of taxa. - 1. The grouping: We all agree that the groups we wish to segregate should be monophyletic, i.e. they must include an ancestral species (known or hypothesized) and all of its descendants (Wiley 1981). This is not an easy concept to test for. When sexually reproductive characters (supposedly highly conserved) like ascus structure are used as synapomorphies, then the segregates encounter wide acceptance among us (see the splitting of Lecidea s. Zahlbr.). When vegetative characters (supposed to be much more open to convergence and parallelism) are used, then the segregated groups are highly controversial, as in the famous splitting of Parmelia. So, should we say that sexual characters should always prevail against vegetative characters? The use of sexual characters is not without problems. Just look at the present segregation trends in cetrarioid genera, where small variations in ascus structure, spore and ascus shapes are used to distinguish small entities. Where are the limits? My point is that we have powerful tools for testing monophyly, those of molecular biology, by sequencing ribosome, and hopefully very soon other genes. I can anticipate that the first (still unpublished) results on the sequencing of the 25S rDNA do not confirm the monophyly of most of the genera segregated from Parmelia s. lat. A possibility would be to require that every new segregation of large groups should be tested for monophyly with molecular methods. 2. The ranking: What should we do with groups which proved to be monophyletic? Pier Luigi is right when he writes "there is no lower limit to the definition of the term monophyletic". Here there will always be a large element of subjectivity, even when molecular methods are used. What we need is: a) sound jugement and wide experience, b) patience: "we should first explore the whole container" before making a sound taxonomic judgement, which often needs more collaboration and less competition, c) withdrawal of the ego (maybe one of the most important problems). Nomenclatural stability should be a high priority, since we cannot go on playing with names without a severe loss of credibility. The problem is that, in taxonomy, both name and author's recognition are connected by nomenclature! I see no other reasons why lichen taxonomists do not like to work at infrageneric levels. One among the possible solutions would be to cite only the basionym, as it is done in Zoology. Nowadays, anybody can validly publish anything, especially in journals that are not reviewed. Publication of new species or new genera should be allowed only in choosed specific journals, as it is the case, I believe, in bacteriology? In any case, editors and reviewers of journals, symposium proceedings etc., as well as the whole lichenological community, should take their responsability when faced to fast-food taxonomy. There is still a long way to go, but the fact that we recognize the problems and that we talk about them is already a good sign. If we want our science to be accepted as basic and modern research we should find new ways of dealing with this important topic. I propose that the IAL could constitute a group, working out propositions, and submitting them to the lichenological community.
Philippe Clerc, Geneva
I liked Nimis' careful presentation of arguments very much. There is not one "good" generic division, but many phylogenetically correct systems are possible and equally acceptable. It is easy to make "good" new genera by just uniting existing ones. Splitting is more tricky, but becomes easier when the segregates are made small enough. Thus, one may say that there is no scientific progress in making a new genus itself. It is only the new arguments leading to this step, which mean scientific progress. If one discovers a good argument to divide a genus into two sister groups, the scientific gain remains the same whether the new groups are treated as subgenera or as genera. The only difference is that in the second case other people are upset, as they have to adapt to new names. Another point concerns the strength of the evidence. If only one character is involved, this is the weakest possible evidence. It could be easily in conflict with the next character to be discovered. Consequently, two questions should be answered when a new genus is to be made: 1. Is there more than one character separating it from its closest relatives, or could any additional character provide an easy argumentation to make the division futile? (If so, users would be happy to keep the current generic arrangement); 2. What are the closest relatives of the new entities? Are they more closely related to each other than to other groups? (If so, treating them as subgenera would not require name changes).
Harrie Sipman, Berlin
The discussion started by Pier Luigi includes widely different scientific problems, as well as covering fields of psychology, sociology, and even subjective believes. Here, I will just touch some points which are mainly connected to our personalities. The naming of things is such an important issue that mankind has been aware of this since the oldest times (Genesis, 2: 19-20, cf. Gen. 1:29-30). Everything has to have a name in order to be or not to be. Naming gives power over the named, and success in the spreading of "our" names gives us some power even on other persons who have to use them. This is why the naming of taxa, events or laws etc. is so important for most people, and not only the fact that authors' names are connected to the taxonomic binomial. Does the zoological system result in fewer new combinations, less splitting and larger monophyletic groups? Yes, some ranks are nomenclaturally more important than others. If I accept or propose a subgenus or a variety this normally does not affect any writer of floras, checklists or general systematic overviews, any curator of herbaria, any teacher etc., but if I describe a new genus, or raise a variety to the species level, our impact on lichenologists is tremendous. The idea of names in current use, and the change of the code in the matter of conservation of names maybe can help us to avoid this, but this has not been the case so far. On the other hand, would we be prepared to let others decide whether our decisions are correct or not? Usually we know the group best, are there any better judges? If we can't agree on some existing persons' taxonomic superiority over the rest of us, it is useless to try to find formalised juries. Systema Ascomycetum or the meeting in Paris some years ago, when we tried to get consensus on the systematics of some lichen groups, can serve as good examples of rejected activities. I think we have to find other ways for a solution.
Jan Eric Mattsson, Lund
I do not agree with most of Nimis' points, for these main reasons: 1) When Nimis argues that lichens are fungi as all other fungi, he implicitly assumes that only - or mostly - sexual characters should be used to separate genera. If this criterion is adopted strictly, it could result in Xanthoria, Caloplaca and Teloschistes being re-united under the same genus. Are we really ready to accept this conclusion ? If not, then the question arises: do lichens have something "special" with respect to other fungi ? 2) "God" did not create genera, genera are names, and names are made, and used, by men. We use names for our practical purposes: uncultured people have a poor vocabulary, cultured people a rich one. The more names we have, the richer our world becomes. Why should lichenologists force themselves to use a poor language? 3) As genera do not exist as such, and as our vocabulary should fit our practical needs, Nimis should explain what is the practical advantage in having a dictionary where some hundreds widely different objects are given the same name, just because they share a few characters considered to be as "more important", such as ascus and spore types. I heard so many generic lumpers in Australia saying: "look, there's a strange Xanthoparmelia"! People around understood much more about the thing than if he had just said "Parmelia". 4) Nomenclatural stability is certainly important. But, fortunately, we no longer call all our beloved, strange beings Lichen. Modern Germans have problems in reading the Nibelungenlied, as many Greeks have problems in reading Homer. Let languages evolve, and among them our language, that of taxonomy !
Louis Le Bois, Venice
Response to Louis Le Bois's remark no.1: I agree that fungi are also (non-lichenized) lichens, but do not think that this implies that only sexual characters are to be used for generic delimitation. In large groups of non-lichenized lichens they are used not only for genus, but also for species delimitation, because there is little else one could use, while the mycelium does not seem to provide many useful characters. In Hyphomycetes many genera are installed without any use of sexual characters. Taxonomy should use any characters that can be properly defined, and objectivity requires that this should be done without a priori weighting. Thus, when lichenized lichens show a wealth of thallus characters, we should use them as best we can. The natural classification of lichens should reflect a phylogenetical process, and is principally different from a classification of physical objects, where a periodical system with few fixed criteria can be very useful. To me, it seems impractical to consistently use comparable criteria for similar ranks, not even in related groups. In Parmeliaceae this would lead to a single genus Parmelia with perhaps 1000 species next to dozens of genera for the 100 species until recently included in Cetraria. It reminds me of putting all insects in a single genus next to Limulus or some other genera of primitive crustaceans. I suggest maintaining supposedly primitive genera, where "basic" characters are variable, like Cetraria s.l. and Omphalodium s.l.
Harrie Sipman, Berlin
Comment to Le Bois' second point - All taxa in one sense are names, but there exists also that reality to which we try to point with these names. Who would claim that that the group of pines (Pinus) with all its diverse species is nothing more than a name? Nature just is highly organized; that's one of the things about it that fascinates us, and it would be unreasonable to expect it to contain species not in related groups. Genera have an obvious reality apart from being mere names.
D. Wright, Fieldbrook, CA
Wright's response to Le Bois exacty reflects what I meant with the sentence "If God created large genera,...etc." and with "The Creation has an inner structure". If genera are not mere names, but a reality, then we can expect that there will be something like large and small genera (e.g. Pinus, Taxus), and our names should fit this reality. Your point makes the deepest core of the whole matter very clear: the splitting of genera just because they are large implies the non-recognition of "genera" as a reality.
Pier Luigi Nimis, Trieste
I disagree, as usual, with Nimis. The point raised by Wright affects any taxonomic rank, not only genera. Yes, we have to deal with a hierarchically structured reality, and this hierarchy is the product of evolution-phylogeny. However, once the structure is covered by a taxonomic hierarchy, the treatment of different taxonomic entities as "genus", "subgenus", "family" etc., is just a matter of naming, and here practical considerations are more than justified, as in the natural evolution of languages. We would not be able to speak about the world if this were not be structured. Our language should cover these structures as a thin, transparent layer. The world is complex, however, and these structures cannot be compressed into a rigid scheme formed by a few hierachically arranged units, those of the same rank so to speak at the same "level". Genera do not exist as such, what exists is the taxonomic structure, and we can use generic names to describe it without worrying about the "existence" of something called "genus".
Louis Le Bois, Venice
Comment on the suggestion that taxonomic analyses should be based on a priori equally weighted characters. - A selection of characters is always made, consciously or unconsciously, because it is impossible to include every character (e.g. "the square root of spore width divided by the depth of the fissures separating areoles"). Any correlation of presumably independent characters is important. Selection of characters and unequal weightings should be made with the greatest care. It is easy to implicitly use a priori judgements of the taxonomy within the studied group when choosing characters, and this must always be avoided.
Lars Fröberg, Lund
Individuals are discrete entities, but their resemblances fall into such a pattern that our mind can assemble them into hierachically arranged groups, such that all entities within one group are more "like" one another than those in other groups. Though the genus is a human concept, it corresponds with something real if similarity discontinuities support its recognition. Otherwise, this would become purely arbitrary. It is my contention that such discontinuities do exist, and that it is the business of systematists to find them. However, the question will still arise as to the hierarchical level of the grouping thus recognized. To answer this question, reference should be made to the degree of dissimilarity - of discontinuity - between this and other groupings. To do this, we need suitable means of measuring similarity between taxa. I have proposed such a measure based on probability. A set of attributes is selected such that no logical or ontogenetic correlation between them seems likely (see the contribution from Fröberg). For each attribute, all possible pairs of alternative values are placed in order of similarity, and for each such pair of values the cumulative probability of that or any more similar pair within the reference population of taxa is assessed. This having been done for all the attributes, then for any given pair of taxa the probabilities are combined over the whole range of attributes, on the assumption that the attributes are uncorrelated. The lower the resulting probability, the more similar are the taxa compared. See an example in Goodall & Marchant (1996, Abstr. Bot. 20, 1: 1-15). It might well be possible to apply this approach to the recognition and separation of genera among lichens.
David W. Goodall, Canberra
One of the main problems of modern generic splittings is that most authors are satisfied with sufficient differences in characters, and that almost all new segregates with new names are proposed without any previous phylogenetic analysis. If we could agree that new taxonomic proposals should be based on some kind of phylogenetic analysis, then much should be gained.
Jan-Eric Mattsson, Lund
I believe that genera exist, contrary to the opinion provocatively expressed by Le Bois. They exist not as objects, but as landmarks in our way of expressing the diverging process of evolution. Populations diverge, become isolated, then reproductively isolated, turn into entities, sharing equal characters from a common ancestor. This is a continuous process, and thinking of this it is hard to accept monotypic genera. In Teloschistaceae, the delimitation of genera is traditionally controversial. Do we have a single supergenus, a few normally acknowledged, less well defined genera, or many smaller genera? Caloplaca is a great example illustrating the evolutionary forces within a group which we can only observe from one point in time and space. Together with my colleague C. Wetmore we have recently worked up a paper on the Gasparriniae for North and Central America. This group cannot be recognized at generic level based upon the character states we have today. It might be accepted as a section or subgenus. However, there are other good natural groups, i.e. some of the black-fruited and semifruticose Caloplacae. A new genus formed from entities earlier included in other genera is in the first place an uncomfortable procedure to learn and adapt to. "Why this splitting?", we may hear from slightly irritated readers. "The old name was good enough, etc." We have all heard this before. But if we have found some new and valuable information in character states, supported by evidence from several strong characters I cannot see the reasons why conservatism or convenience should prefer to hang on to something which was based upon irrelevant data from the good old days. Some of the genera from these days were good of course and still stand. Many were actually very bad. When I started my career, Alectoria, Lecidea, Lecanora and Parmelia included everything what we today have divided into a very large number of smaller groups. I guess that nobody would agree that these four genera from the early 1970's represent monophyletic groups. Where will the descriptive process end? Many of us contributed to move the snowball down hills, and people may ask how will it all end: "as an avalanche - I guess many would believe - obliterating the idyllic small ancient village down there, in the Acharian valley of the past". We can hear the most irritated claim that we will have nothing but genera left in the Parmeliaceae. Have we been moving the snowball in the right direction? Yes, I definitely believe so. In the other direction, i.e. "uphill", back to the good old days, would be against all natural forces and impossible at the end. We cannot state that everything which is published as new on the generic level is great and outstanding. But who can decide what is right or wrong here? Nimis was provocatively challenging us with the "Creator" talk. Who can ever decide what is a good genus or not, or who is our "Lord"? The "Creators" of Systema Ascomycetum, perhaps? Where we can actually read what is good or not, or accepted or not. Life on earth was not made in seven days but merely during a period of 1.5 billions years. Psiloparmelia or Flavopunctelia might still be great genera. The enormous number of new groups which have been separated during the last decades have taught us a lesson, perhaps to think twice before doing anything. Groups based upon one character don't seem particularly good. Even using 2 or 3 not very convincing characters in the definition of new genera doesn't help us much. The so-called genus Gasparrinia can be defined on 3 not very good characters. That's all. So I am not going to do it. If somebody would separate the usnic acid containing Caloplaca as Flavocaloplaca, the vulpinic acid containing Rhizocarpon as Vulporhizocarpon or the bluegreen containing Peltigera as Cyanopeltigera, these groups would probably not be allowed into the "Lords" system. But who knows... This might pass even quicker than the Seven Days of Genesis.
Ingvar Kärnefelt, Lund
I am an old man now, but I will say a few things about taxonomy anyway. First, no reputable taxonomist would describe a new taxon based on only one character, whatever the character. The ascus tip (used by some to describe new families) has proved to be variable within one genus! Then why describe new families based on this single character? Have they looked at all species in the genus/family to understand the variability? No. Only the type species and then usually only one specimen. This is no way to study evolution/taxonomy. What is going on with systematics? Do we have to blindly follow anything that is published? Many do, but the best do not, they evaluate each one, as my old master taught me. As a result of this teaching, I was one of the last in North America to still believe in Parmelia. Granted, there are several good genera there but NOT 90!!!! Do we have to divide genera until there are only 2-5 species per genus? Many of these genera have been erected either on a few characters or because "If I don't do it, someone else will" (as a famous lichenologist once said). Another reason for describing new genera is for the "notoriety" of having your name cited every time someone types a label (on some it even takes 4-5 authors to make up their minds on a nomenclatural change; some of the new author citations are too long for any label!). Is this responsible taxonomy or taxonomic inflation? It is time to remove author citations! Perhaps I have lived too long. Maybe everything has passed by. I think I received an excellent training on being a thoughtful taxonomist, but maybe that is not in fashion now. Maybe now the main goal is to divide and describe. I am sorry for taxonomy if that is true.
Cliff Wetmore, St. Paul
Many thanks to all contributors. This discussion was started and finished in less than 30 days using LICHENS-L, Cliff Smith's listerver (see next section and try it...you can get most of the original texts, who did not fall under the usual editorial castigations). Of course, the discussion does not stop here: it will go on in Montreal, London, and, as far as I know, it is already going on in many other places, including several bars and pubs near botanical institutions. In the next issue of the Newsletter, however, only a very limited space will be available for this topic, and this will be mainly reserved for contributions by colleagues without access to e-mail. New proposals for the next Forum discussion are welcome; this time, priority will be given to topics outside the field of taxonomy s.str.
The Editor