1. Future of lichen bioindication
2. Phycobiont versus photobiont
3. Specimen citation in WWW

Topic 1: The future of bioindication with lichens, by Ch. Van Haluwyn

More or less twenty years ago, Van Haluwyn (Cons. Scient., Univ. Lille 2, 1978), Lerond (Act. Mus. Rouen, 1-2, 1981) and Van Haluwyn & Lerond (S.R.E.T.I.E., 1986; Coll. Phytosoc., XV, 1987; Cryptogamie, Bryol.-Lichénol., 9, 1988) proposed a phytosociological approach to air quality monitoring with lichens. As stated by Iserantant & De Sloover (Mém. Soc. Roy. Bot. Belgique, 7, 1976), plant records (relevés) of stands give more information than the registration of single species. For this purpose, we applied the classical procedures of the French-Swiss school (Braun-Blanquet). We thought, like Wirth (Bibl. Lichenol., 30, 1988), that a phytosociological relevé contains a great deal of information, and may be analyzed in different ways as it includes a complete set of qualitative (floristic) and quantitative data. A statistical treatment of more than 600 relevés distributed in half the north of France revealed seven groups of epiphytic species distributed along a hierarchical gradient of impoverishment, from associations through alliances, orders and classes. There is sufficient evidence that the depauperation of lichen communities is related to air quality (not a summed effect of a variety of pollutants, but the result of antagonist and synergic interactions among pollutants); hence, we proposed an eco-diagnostic scale of environmental quality using 30 epiphytic lichens, choosen inside the seven groups of impoverishment. The maps produced with this scale in northern France agreed well not only with the location of the main emitting sources and the prevailing winds, but also with different environmental conditions (acidification, eutrophication, etc.), and our scale seems to reflect the global quality of the environment, providing a frame for a qualitative estimation of lichen diversity, in agreement with Nimis' concepts (see e.g. Cislaghi & Nimis, Nature, 387, 1997). This is a logical evolution of the qualitative method established by Hawksworth & Rose (Nature, 227, 1970). Sulphur dioxide alone is no longer the major factor affecting lichen distribution in urban and industrial areas. The total epiphytic vegetation is a more realistic marker of the sum of environmental conditions. The lack of English translations led to non-recognition and/or misunderstanding of our approach. Some controversial points are listed below:

- the sample surface of the relevè is defined on a rather subjective basis (concept of minimal area); although this appears as one of the main weakpoints of our method, there would be no problem in standardazing the sampling strategy;

- the statistical treatment of the relevés is problematic: however, good results are obtained by using factorial correspondence analysis (see Wolseley & Pryor, Lichenologist, 1999). Syntaxonomy is by no means fundamental for the application of this method. Swiss, Italian and German lichenologists have simplified the quantitative method proposed by Leblanc & De Sloover (IAP), and a guideline has been proposed for measurement and evaluation of phytotoxic effects of ambient air pollutants (Wirth, VDI 3799, 1995). In my opinion, the two methods are not in contrast: they express lichen diversity in a quantitative or qualitative way. The examined unit for VDI is a number (sum of frequencies), while in the phytosociological approach this is a relevé (list of species with cover-abundance estimates); the final result of VDI is a series of classes, that of the phytosociological approach is a sequence of syntaxa (groups of species with similar ecological characters, such as acidification, eutrophication, toxitolerance, climatic factors). This, in my opinion, is an advantage with respect to the quantitative method. For a general acceptance of monitoring methods based on lichens the results should not only be calibrated against pollution data, but also with other data, such agricultural pratices, urbanization, traffic density, health standards...Which method is the best? This is an open question. I would respond: it is a problem of feeling. It is dangerous to impose a single method: lichens are living organisms (lichenologists also!): we must take into account both regional features of lichen vegetation and environmental conditions (nature and number of phorophytes, climatic factors) and also of the lichenologist's feelings. An integrated approach where both methods are used jointly is probably the most appropriate (see the numerous papers by Nimis for a review). On the other hand, standardization of procedures cannot but enhance the success of lichen bioindication. These observations need, perhaps, a discussion.

The future of bioindication with lichens using phytosociological methods will always be clouded by the absence of a causal agent in the analysis. The relationship between lichen presence/absence and air quality in such work is correlative only. This is why I prefer to use chemical content indication instead. Elevated concentrations of anthropogenic chemical elements are directly measurable and can be related to emissions as well as presence/absence of sensitive species. It only requires that samples be taken and analyzed in the laboratory. The relationship between concentrations of chemical elements in lichens and air quality, however, must be researched more, in order to separate it from other sources of elements in tissues.

Phytosociology s. str. has two main weak points, one is subjectivity in the sampling strategy, the other is what they call syntaxonomy (too a long story for this short note...). In my opinion, phytosociologists are among the worst enemies of phytosociology: instead of modifying just slightly their method without basically changing its original, powerful approach, they often tend to hide subjectivity into a plethora of pseudo-epistemological considerations dressed in dense clouds of "Latinorum", using a jargon which is totally undigestable to non-initiates. This is a pity. The Swiss-Italian-German approach to lichen bioindication mentioned by van Haluwyn, on the contrary, adopts a strict sampling strategy for measuring lichen biodiversity. Contrary to the opinon of Bennett, the problem of correlating these measures with pollution data is basically the same as for metals. Ecology is a science of complexity and multi-causality: simple one-causal explanations do not bring too far. The French approach has little to do with the negative sides of phytosociology, and is basically interesting, but it maintains a high degree of subjectivity in the sampling stage, and its "local" character makes any generalization quite difficult: a complete phytosociological analysis must be repeated every time in different areas, obtaining different groups of species, and different indicator species. On the contrary, the Swiss-Italian-German approach (...VDI being quite odd, is there a shorter name?) brings just to a simple measure of biodiversity, a single number. This is a purely biological datum (parenthesis for some colleagues: datum is singular, data is plural: the data is sounds horrible), which is interesting by itself, irrespectively of its possible correlations with air pollution or with other factors. In most cases, however, correlations with pollution were found, albeit different in different areas. At the end, I do not see any relevant differences between the two methods: the frequency data deriving from the Swiss-Italian-German approach can be (almost always are!) organized into a matrix of species and relevès, which can be processed exactly as a matrix of phytosociological relevès. The two approches should not be used together, they should be merged into a single methodology.

I didn't mean to imply that studies with chemical elements would elucidate single causality explanations, and so I thank Le Bois for his comment. In fact, I am a strong believer in multi-causality because that is the nature of our world and the field science we do (as opposed to laboratory science). My papers are becoming more and more multivariate with each new study. But why do we do bioindication studies in the first place? What is their basic premise? We do them because we theorize (suspect) that air pollution is killing off lichens. It may even be killing off species, but that's harder to prove. I work closely with pollution regulators, who must have direct causality proven in order to do something about the pollution in the first place. Some of us may do bioindication studies for other reasons, e.g. biological interest, but I do them to generate data and evidence that air pollution is not good for plants, and hence not good for us, in the hopes that my data will help doing something about it. There - I admit it - I'm a scientist with a purpose. So I think we should be doing bioindication studies to demonstrate causality and effects on floras, if possible. And if Le Bois or anyone else can come up with a way to merge all this into a single methodology I will join in eagerly.

I totally agree with van Haluwyn, who says that a combination of a quantitave and a qualitative method will bring the best results in describing the state of the environment using epiphytic lichens. On a small scale such a method exists, as in the Ruhr-area: several investigations took place using the approach by Rabe & Wiegel (1980), the so-called LUGI (Luftguete-Index), which uses species frequencies (combined with a species-concerned factor), but also ranges the values in classes based on associations of differently sensitive taxa. The same approach is followed by the HTI (Häufigkeit-Toxitoleranz-Index, Kricke 1998), taking into account all species found on a sampling site with values ranked according to characteristic species communities. This region has a very impoverished lichen flora, due to past high levels of air pollution, and now the situation has improved: a method addressing both to the abundance and the composition of lichen vegetation seems to bring good results. The "original" IAP by DeSloover & LeBlanc considers also the association of lichens: they used a weight factor for each species, based on "adding together the number of species of epiphytes escorting it at a station and then taking the average of the sums for all stations where that species was present". I think that this method is quite as good as other techniques.

In my opinion the view of van Haluwyn is basically correct: qualitative and quantitave methods should be used jointly to describe the general state of the environment. Contrary to Le Bois, I do not like much the fact that biologists should always reduce their (biological) data to numbers. IAP values are numbers, but the same number can arise from very different lichen assemblages, each one with its own ecological meaning. An example: with the Swiss-Italian-German method one can have IAP=10 with a single species present, in one case with Parmelia caperata, in another with Amandinea punctata. I shall say that air quality is almost the same, but, ecologically, the two environments are quite different. An integrated combination of both approaches is probably the best for describing the environment (not only air quality). Of course, politicians, land planners, pollution regulators etc. need data which are easy to be read (such as zonal-maps with short sentences clearly explaining the situation). Two further remarks: 1) I started my career as a phytosociologists of higher plants but, unfortunately, I now totally agree with Le Bois: phytosociologists are among the worst enemies of phytosociology. 2) The Q-factor in the original IAP formula by LeBlanc and De Sloover is not subjective, as stated by someone (it's just a count), but it is strongly affected by the sampling strategy: the number of companion species will depend on the total number of sampled sites, their distribution in the survey area, etc.

A recent message from an American colleague complained that in N America biomonitoring is often considered as a quaint European anomaly, while I always thought that the scarcity of such studies there is a quaint American anomaly. This, in my opinion, is due to two widespread (and wrong) attitudes of biologists in selling their methods: 1) Biomonitoring estimates air quality. When asked to define air quality, biologists start with a deluge of more or less eco-philosophical explanations, which can be summarized as follows: air quality is that thing which is measured by indicators of air quality. This is odd. 2) Biomonitors are cheap instruments to monitor air pollution. Pollution is defined "operationally" (an important term coined by P.W. Bridgman, trivial in Physics but almost unknown in Biology, details upon request), in terms of instrumentally measured concentrations of given substances. To do this, we need instruments, not organisms. Biomonitors, by definition, do not measure pollution, they measure its effects on pollution-reactive organisms: biomonitoring is a branch of Biology. Lichenologists can be proud of an absolute record in publications demonstrating that lichens can be effectively used as monitors...but of what? Air pollution? Air quality? Air purity? In my opinion, lichen data just provide an estimate of deviations from normal conditions of pollution-reactive organisms. Nothing less, nothing more. Different scales for interpreting our data will be needed, because "normality" is different in different situations. Instead of insisting on the calibration of lichen data against recording gauges, we should give more stress to the quantification of background conditions in different situations (metals in areas with different lithology, biodiversity in bioclimatically different areas). If I could (I can't), I would delete the ambiguous terms air quality and air purity (which sounds still worse) from any lichenological dictionary: as long as we use our beloved, historically so important term I.A.P., we'll be on the wrong way. Even when presented as they really are, lichen data will still have a good value on the pollution monitoring market: they might be more or less correlated with instrumental pollution data: generally they do, but they have an interest in themselves, as biological markers of ecosystem disturbance, without being either measurements of air pollution, or estimates of an all too vaguely defined air quality.

There are two ways in which lichens are used to study air pollution and its effects. One has to do with physiology and ecology (community assemblages, etc.), the other with contaminant concentrations in the thalli. I have been using the term contaminant landscapes to describe the latter. This gets around some of the objections raised by Nimis, as it does not involve direct use of the term biomonitor. It does reflect the fact that what we are looking at really is contaminants "on the ground" (or in the lichen), rather than in the air.

When biomonitoring is defined, like I suggested, as "estimates of deviations from normal conditions of pollution-reactive organisms" there is no basic difference between accumulators and indicators.

Phytosociology has proved itself as a useful tool to identify associations of species in their natural state and, as van Haluwyn says, provides both qualitative and quantitative data. However, lichen communities of habitats where environmental conditions are changing are species-poor and do not conform to described phytosociological taxa (see Wolseley & Pryor, 1999). Sensitive species disappear, and are replaced by thoses tolerant of the altered conditions. Such communites may contain rapid invaders of the new conditions, and long-lived specimens of species not able to reproduce in the new conditions. A recent investigation of Hypogymnia physodes in Ekaterinberg (Russia) showed that soredia production was dramatically lowered in polluted areas, with changes in the population structure (Mikhailova, 1998). Conversely in pasture on the edge of an oak wood in west Wales (UK) the twigs supported abundant small Usnea initials, but few mature specimens within an atypical Physcietum adscendentis, suggesting that environmental or competitive conditions acted to depress growth in a site 150 m away from the propagule supply. In contrast to the range of biological effects in very local conditions, atmospheric recording data are only collected in scattered sites many miles apart. The modelling of movements of pollutants in the atmosphere is still being investigated. In 1989 the 59 Precipitation Monitoring sites in the UK were reduced to 32, the cuts being mainly in 'clean' areas which may be surrounded by arable or highly fertilised leys where conditions are changing rapidly. Data is collected weekly at these stations so that weekly averages are recorded, not pollution events. Continuous monitoring is mainly restricted to urban and built up-areas, yet the threat to lichen communities may well be in our agricultural areas. Biological monitoring cannot replace atmospheric or other monitoring, but it allows assessment of environmental conditions at each recording station. In the vicinity of La Spezia, Nimis was able to interpret data from 76 biological recording sites using information from 7 continuous atmospheric recording stations. If we accuse phytosociologists of being fixed in their approach, then we must also allow that any selection of species as indicators becomes biased towards already established conditions. e.g. Francis Rose has changed and refined the indicators of ancient woodlands over 20 years, but he also shows that we need a western, northern and eastern list in the UK. There is a danger that once lists are published and in use they become fixed, and intuition has no place. Phytosociological sampling provides a whole data set that we can reinvestigate at a later date to detect environmental conditions which we may not have thought of. Twenty years ago we were concerned with acidification, but now we are monitoring the return of species to our inner cities, and in the country the replacement of whole lichen communities with the Buellietum punctatae or, worse, with green algae. Over what areas of the world is this happening?

In the international literature the following distinction is often found: a) "passive" biomonitors, i.e. those already present in situ (e.g. lichens on trees), b) "active" biomonitors, placed in the survey area by the operator (e.g. lichen transplants). In my opinion, the terms passive and active, in this context, are not only unelegant, but also quite confusing, e.g. with active and passive processes of metal uptake. This terminology, which is already entering in the legislation of some countries, should be abandoned altogether. An alternative could be something like: autochtonous and allochtonous, but probably there are less exotic terms in English. By the way, the active-passive terminology is often used in expressions like passive biomonitoring, which, of course, cannot be translated into autochtonous biomonitoring. In any case, such studies are neither passive (why?) nor autochtonous. Any suggestion?

How about: naturally occuring biomonitors vs. transplants? I am not in favor of developing more jargon. Plain speaking is best.

I agree with Nimis about the confusion arising from the terms active and passive biomonitors, which, to tell the truth, is mostly caused by scientists using lichens as bioaccumulators. In my opinion the term active should refer to something actively indicating a process. In the case of bioaccumulation, lichens do not react directly (i.e. they hardly die because of a too high uptake of a certain metal), but accumulate more or less passively. As suggested by K. Enns (see below), when transplants are used, in order to avoid further confusion, terms such as native or transplanted are probably clearer.

The passive-active terminology is a problem, it is true. I use native and transplanted Peltigera aphthosa. I refer to lichen colonies in situ as native lichens, while transplanted lichens are called, not surprisingly, transplants. The latter do not have a lengthy history of exposure, being subjected to transplant shock and other problems associated with an overnight trip to a slope next to a lead smelter or an old pulp mill that has been rebuilt. The natives, however, have a long history of declining exposure to emissions. I am interested in the concentrations of metals and sulphur in their tissues, and in the biochemical fate and toxic history of these elements. The two types of bioindicators are not comparable. This is why I use other things, like sulphur organic to inorganic ratios in two-three year old conifer tissues, symptoms of injury in vegetation, species compositional changes over time, and concentrations in soils at the rooting layer and in the litter-fibre-humus layer. I look also for linkages with invertebrates (bioaccumulation) and for toxicity fates, to sketch an air quality profile for areas impinged by heavy metals, sulphates, nitrates, ozone, etc. I cannot resist adding that here in Canada passive monitors are little disks which we use (similar to sulphation plates) to measure ambient sulphur dioxide and NO_x.

I fully agree that these terms, especially when relevant to accumulation of substances, are absolutely confusing. When explaining the different strategies of biomonitoring to other persons (e.g. students), somebody always thinks that the operator is passive, as he just wanders around, listing lichens growing on trees. In the case of active biomonitoring, the scientist is active as he has to find good growing lichens on one place, and transfer them to well-elected places anywhere else. Thus, I support the suggestion that the term bioindication with native lichens should replace the current passive bioindication with lichens.

A very final remark: do really lichens actively accumulate metals, or is this just a matter of passive deposition? If the latter is true, why to use lichens and not some artificial material with good absorbing properties? And, if this is true, is the term accumulation correct? Accumulatio, in Latin, just means heaping up, which works both for passive and active processes. Barnacles actively accumulate metals from water by filtering several liters every day, but lichens also accumulate metal particles deposited upon them every day. The two processes are basically different (one is active, the other is passive), and require two different terms (no new jargon, just a sound definition of concepts). - Maybe, when I'll be old enough, I'll write a paper in the I.P.N (International Philosophical Newsletter) about the confusion created in Biology by the uncritical use of the term accumulation: in my opinion, physiological (active) mechanisms of metal accumulation in lichens might have been largely overestimated, most of their metal content being due to passive incorporation into the thalli, which means that lichens are as good as any other deposimeter. Is this really so a terrible threat for applied lichenology? And, if yes, why? To elucidate this problem we would need another Forum, some further research...and/but perhaps just a good critical review of the existing literature.

Topic 2: Phycobiont versus photobiont, by W. B. Sanders

I would like to suggest that the term phycobiont be restored to common use. It was replaced by the term photobiont, based on the view that the cyanobacteria, which are photoautotrophic symbionts in many lichens, "are not algae per se" and therefore not included in the concept of phycobiont [Phykos, seaweed] (Ahmadjian, Int. Lich. Newsl., 19, 1982). Most subsequent publications (including my own) conformed to this new terminology. However, on closer consideration, I believe the adoption of photobiont for phycobiont is not well justified, and has certain disadvantages. What are algae, and what is the basis for excluding the cyanobacteria - the blue green algae - from this concept? Like many terms in biology, algae can't be perfectly defined. It is based on "lifestyle", not on phylogeny. Most organisms we call algae are photosynthetic, prefer aquatic or marine habitats, and share a rather nebulous "grade of evolution," which in this particular case might just be a fancy way of saying that the embryophytes are excluded. This is far from precise, but the term has endured simply because it is of some use. It should be noted that phycology courses continue to include the cyanobacteria/blue-green algae, as do the most recent textbooks (e.g. Van Den Hoek et al., 1995: An Introduction to Phycology. Cambridge Univ. Press, UK). Nowadays, though, we are so preoccupied with phylogeny that sometimes we think all categorizations of living things have to reflect evolutionary ancestry. Certainly those we use for systematic purposes ought to. However, terms such as algae will never be of systematic use. While it is significant that the cyanobacteria are prokaryotes, classified in the Kingdom Monera, why should this disqualify them from being algae? Excluding them certainly won't make the remainder monophyletic or even paraphyletic. That is not the point of the term algae, and can't ever be. So, what value is there to re-delimiting the algae as eukaryotes? I don't think there is any, and since phycologists continue to treat the cyanobacteria within the concept of algae in their texts and courses, lichenologists ought to follow suit. Thus, phycobiont should be appropriate for both green and blue-green algal symbionts, as originally proposed by Scott (Nature, 179, 1957). This usage has been maintained by some lichen researchers (e.g.Tschermak-Woess, in: Galun: Handbook of Lichenology, I, 1989), and is still recognized in the latest edition of the Dictionary of the Fungi (Hawksworth et al.: Dictionary of the Fungi. 8^th ed. CAB Int., Wallingford, UK). As compared to phycobiont (in its original inclusive sense), the term photobiont has some disadvantages. One is that it is rather less specific. Since photobiont merely indicates that the symbiont is a photoautotrophic one, it could equally well apply to, say, the host plant in a mycorrhizal relationship. A second point is that photobiont does not logically contrast with mycobiont. If we refer to the algal symbiont by its mode of nutrition, then one might expect the fungus to be treated similarly, with an unwieldy parallel term such as heterotrophobiont. Of course, photobiont is now in current use, and many of us are used to it. But there is also value in keeping terminology as simple and logical as possible. It gives newcomers and those in related fields easier access to the literature, facilitating comprehension and the contribution of fresh viewpoints.

Keep both. Use each when appropriate. No point in trying to lock in one term or the other.

William makes several good points. My impression, however, is that much of the current lichen literature does not use the term photobiont in a way which is exactly interchangeable with phycobiont. Rather, the former is used when the author wishes to emphasize the photosynthetic role of the green alga or the cyanobacterium, while phyco- and cyanobiont are used in other cases, to emphasize the distinction between eukaryote and prokaryote, which also facilitates comprehension. Whatever one wishes to call them, Chlorophyta and Cyanobacteria are quite different critters.

I have to agree with many of William's arguments regarding the use of phycobiont vs. photobiont in reference to the photosynthetic symbiont in lichens. I believe, however, that the damage has already been done, and readopting phycobiont for photobiont would destabilize our terminology too severely. Photobiont is still easily explained to newcomers to the field as a short form of "photosynthetic symbiont", so it works. As much as I lament the loss of the good old term "blue-green algae" (and I still use it to some extent), I don't want to backtrack and cause new confusions. But, if the lichenological community agrees that phycobiont should be reinstated for the reasons so well expressed by William, then please tell me about it soon. I use the term photobiont about a thousand times in Lichens of North America, which is due at the publisher in a few weeks!

Recently I had a few somewhat complicated discussions with an undergraduate student and lichenology-beginner just circling around the simple question "What is a lichen?", and therefore I agree that we should keep terminology as logic and simple as possible. I'm working with blue-green lichens, and I normally use the term cyanobiont. However, for me it wouldn't cause problems to have a single term for any given photoautotrophic symbiont within a lichen thallus if this is not used in a phylogenetic context. Looking at the phylogeny of the phycobiont s.lat. there are already terms like Cyanobacteria, and Green Algae including subordinated taxa, which are suitable to be addressed to whenever phylogeny needs to be stressed.

Topic 3: Citation of specimens in monographs and the WWW (continuation of the Forum started in I.A.L. Newsl., 30, 2)

I would like to come back to an issue already treated in a Forum-discussion. Many journals do not accept long lists of examined specimens, even in monographs. Such data, which may contain information of interest (e.g. more details on localities, substrate, accompanying species), are often available from the authors of monographs. It would be more convenient for someone asking for distributional data if these could be available on the Web. Is it a feasible solution to submit specimen data used in monographs to an on-line accessible database? It would be analogous to the submission of sequences to the GeneBank database, which is often a prerequisite for publication of molecular results. An on-line database could be dedicated to this purpose, and Authors should submit their data in a format which will not require much work for inclusion into the database. Editors of journals, however, should agree on this system. Any remarks?

I agree with Martin on making complete specimen lists available on the Web. I have complete specimen lists for all of the Caloplaca species I have treated. I wonder how many others have similar databases. There are several problems with the idea, but none too big. One is where to put the lists. My Web page is not big enough for all the lists as well as the keys, maps and photos, and I suppose this is the same for others. It probably should be in one central place.Then there is the problem of standardization of format, and the database system to be used. What about updates by the person submitting the lists? What about the longevity of the lists? Are they there for years or what? Someone will have to maintain the central system too. Maybe updates would not be desired if the cited-specimens database is used as a electronic supplementary to monographs. Each entry in such a database would then refer to a paper-bound publication and should remain stable. This also anwers the question of longevity. If there are additional specimens to be included in the database, this should be by a separate submission, accompanied by a reference to another paper-bound publication.

Maybe updates would not be desired if the cited-specimens database is used as a supplementary to monographs, where each entry should refer to a paper-bound publication where this information is just referred to, and should remain stable. This also anwers the question of longevity. If there are additional specimens to be included in the database, this should be by a separate submission, and accompanied by a reference to another paper-bound publication.

This is not quite what I was thinking. For some species I have several hundred specimens in my database, but I could only publish, say, 60. The rest, along with others I have seen since the publication, will probably never see print. To be most useful, the database of specimens should include all material that has been confirmed by the monographer. If the database only includes the specimens in a paper, there is no need for the Web database.

The idea to make such specimen lists available via the Web has my support. I have no lists to offer myself, but there might be space on the server here at the Botanical Museum, and I could keep an eye on them for the coming ten years. If the lists will be provided as text files, arranged after species and country, I could place them on the server as htm-documents and add clickable lists of species and countries. This would make it easy to find information on a certain species from a certain country. I guess that such a structure would be sufficient for most requests. Information from other databases could be re-arranged, accordingly, by me, and transformed into text files.

Why clutter up the electronic airways with cited specimen details. Would it not be better to lodge a hard copy of cited specimens in a national herbarium library where the bulk or many of the speciments consulted are housed? Thatway follow-up researchers would have access to the data.

I find the idea of having available on the Web an electronic databank filled with published and unpublished cited specimens very good indeed. As we all know, Editors are always reluctant to publish such data, opting for maps instead. Maps, however, do not contain all the information, and it would be great if we could refer to a cited specimen data bank available on the Web. The analogy with evolutionary trees and the DNA data bank is perfect. The biggest problem is that not everybody has access to the Web (only 2% of the World population). I don't know what is the proportion of people interested in cited specimens with access to the Web, but I guess they are not the majority! In my opinion, such a data bank should fulfil some conditions: 1) Standardization of format and data, 2) Each submission should be linked with a paperbound publication, 3) For each data set (one record for one species) it should be clear who submitted the data and what paperbound publication is linked with, 4) Updating of submissions should not be allowed: a new submission linked to a new paperbound publication would be necessary, 5) New data from specimens seen after the publication of the paper could be entered in the database as far as the same taxonomic concept is applied as it was in the publication. Of course, only the author of the publication should be allowed to enter such new data. I hope the discussion will go on adding new ideas.

The idea sounds utopian, but for how long will the data be accessible? Think how many of us have access to the miles of large format tape recorder discs, the tons of information on old large floppy discs, or years of metereological data from Arctic research programmes on punched cards from the 70's. The Web and e-mail, in my view, are wonderful for short term discussion, information access etc., but as a long-term accessible repository this system has still to prove itself (and it is not free: someone has to maintain all these data bases and links, as well as the technology). What happens when a person dies? You have only to try, and contact Web sites that were operational a year ago, to find that 30% have changed, or are no longer available.

Imagine someone needs to prepare a distribution map of Toninia sedifolia in Québec. Of course, you expect that the monograph by Timdal (Opera Bot., 110, 1991) will give you some information. On page 97, you find some data for the presence of the species there: BM, UPS, US: 1. Seeing the long list of T. sedifolia specimens, you will understand why author and editor agreed on some kind of abbreviation. However, you still want to make the appropriate dot for this specimen in your map. Now you can phone or e-mail Timdal for more information. If he cannot be contacted, you may wait for an answer by the curators of the referenced herbaria, or you may wonder whether and where there is a hard copy list with the full information. If time can be saved here, it might be used better, e.g. for more fieldwork.

I am actually planning to put all specimens I have examined of Toninia on the WWW, in a system similar to that now working for the Norwegian Lichen Database and Recent Literature on Lichens (http://www.toyen.uio.no/botanisk/lavhb.htm). This database will include the specimens on which the monograph is based, and all those I have examined after the monograph was completed. I will not restrict myself to not make corrections or additions, so the dataset will not reflect accurately the data presented in the monograph.

To avoid the problems rightly addressed by Richardson we should propose Grube's good idea to a much wider Forum (e.g. International Botanical-Zoological Congresses). Why should something like that be possible for Gene-banks, and not for distributional data? A dot on a map is often much more difficult and costly to get than a gene sequence.