The Challenges of Taxonomy

commentary Challenges for taxonomy The discipline will have to reinvent itself if it is to survive and flourish.

One reason is that taxonomists lack clearly H. Charles J. Godfray achievable goals that are both realistic and rel- Taxonomy, the classification of living evant. Of course it would be great to describe things, has its origins in ancient Greece and every species of organism on Earth, but we are in its modern form dates back nearly still monumentally uncertain as to how many 250 years, to when Linnaeus introduced the species there are (probably somewhere binomial classification still used today. Lin- between 4 million and 10 million); this goal is naeus, of course, hugely underestimated the just not realistic at present. There are various number of plants and animals on Earth. As projects aimed at listing, for example, all the subsequent workers began to describe more valid described species of animal in Europe, and more species, often in ignorance of each or butterflies on Earth (see Box 1, overleaf). others’ work, the resulting confusion and These aims are eminently achievable and very chaos threatened to destroy the whole enter- worthwhile, but the results are like raw, un- prise while still in its infancy. In today’s annotated DNA sequences: unexciting and of jargon, we might call this the first bioinfor- relatively little value in themselves to non- matics crisis. Using the tools then available, specialists. Taxonomists need to agree on nineteenth-century taxonomists solved this deliverable projects that will receive wide sup- crisis in a brilliant way that has served the port across the biological and environmental subject well since then. They invented a sciences, and attract public interest. complex set of rules that determine how a A second problem is part of the legacy of species should be named and associated more than 200 years of systematics. Many with a type specimen; how generic and high- taxonomists spend most of their career try- er taxonomic categories should be handled; ing to interpret the work of nineteenth- and how conflicts over the application of century systematicists: deconstructing names should be resolved. All these rules their often inadequate published descrip- From paper to screen: is it time for taxonomy to revolved around publications in books and tions, or scouring the world’s museums for break with tradition and unify on the Internet? scientific journals, and their descendants type material that is often in very poor con- form the current codes of zoological and dition. A depressing fraction of published being a scientist at this particular time in biological nomenclature. systematic research concerns these issues. history is the vast amount of information But today much of taxonomy is perceived In some taxonomic groups the past acts as a that is available, essentially free, via one’s to be facing a new crisis — a lack of prestige dead weight on the subject, the complex desktop computer. I can download the and resources that is crippling the continu- synonymy and scattered type material sequences of millions of genes, the positions ing cataloguing of biodiversity. In the United deterring anyone from attempting a of countless stars. Yet, with a few wonderful Kingdom, a Parliamentary Select Commit- modern revision. As Frank-Thorsten Krell exceptions, the quantity of taxonomic infor-

tee is currently conducting an enquiry into pointed out in Correspondence (Nature mation available on the web is pitiful, and MUSEUM, LONDON HISTORY NATURAL the health of the subject for the second time 415, 957; 2002), “original descriptions have what is present (typically simple lists) is of in 10 years, and similar concerns are being to be referred to for ever, independent of the little use to non-taxonomists. But surely tax- expressed around the world. In this article paper’s quality”. onomy is made for the web: it is an informa- I shall first explore why descriptive taxono- The problems do not always lie in the past. tion-rich subject, often requiring copious my is in such straits (in contrast, its sister Even today, many species are being described illustrations. At present, the output of much subject, phylogenetic taxonomy, is flourish- poorly in isolated publications, with no taxonomy is expensive printed mono- ing). Then, after this essentially negative attempt to relate a new taxon to existing graphs, or papers in low-circulation jour- exercise, I will argue that taxonomy can species and classifications. Many of these nals available only in specialized libraries. prosper again, but only if it reinvents itself as ‘new’ species will have been described before, These are not attractive ‘deliverables’ for a twenty-first-century information science. so sorting out the mess will be the headache of major research funders. It needs to adopt some of the solutions that the next generation of taxonomists. It is not molecular biologists have developed to cope surprising if funding bodies view much of Two models of taxonomy with the second bioinformatics crisis: the what taxonomists do as poor value for money. The taxonomy of a group of organisms does huge explosion of sequence, genomic, pro- One of the astonishing things about not reside in a single publication or a single teomic and other molecular data. institution, but instead is an ill-defined integral of the accumulated literature on that The problem group. The literature is bound together and Why can’t descriptive taxonomy attract his discipline is cross-references itself using the venerable large-scale funds in the same way as other big made for the web: rules of taxonomy encapsulated in the codes. programmes like the Human Genome Pro- T But this is not the only way to organize a tax- ject or the Sloan Digital Sky Survey? All three it is information-rich onomy. The taxonomy of a particular group projects are enabling science: not in them- could reside in one place and be adminis- selves generating new ideas or testing and often requires tered by a single organization. It could be hypotheses, but allowing many new areas of self-contained and require reference to no research to be opened up. copious illustrations. other sources.

My main argument is that to address the shall sketch one possible way a unitary tax- type material that underpin distributed tax- problems outlined above, and for taxonomy onomy might be achieved. I am not a profes- onomies do require administration, which is to flourish now and in the future, it has to sional taxonomist and am under no illusion currently undertaken by our great museums move from the first to the second model: that what follows will be the best or even a and herbaria. Nearly all these organizations from having a distributed to a unitary orga- viable model, but I hope it will bring out the are enthusiastically embracing modern web nization. Such a massive task could only be issues involved. technologies. Hosting web revisions is some- accomplished group by group, as resources thing I see as a logical extension of their moves became available. I believe a number of A unitary taxonomy towards becoming, in part, modern informa- things would then follow. First, the only Introduce as a formal taxonomic procedure tion storehouses. It is absolutely clear, how- logical way to organize a unitary taxonomy the ‘first web revision’. This would be a revi- ever, that they need more money in order to and to make it widely available is on the web. sion of a major group of organisms to a stan- do this. They might also undertake the intel- The web is currently used, if used at all, as an dard decided on by the International Com- lectual administration of the web revision — adjunct to the distributed, printed taxono- mission on Zoological Nomenclature, or the the refereeing and editing — although they my, but I think it should replace it. Second, International Botanical Congress, or equiva- would probably devolve this to committees the core of taxonomy is a description of each lent body (let’s just call it the international drawn from a wider constituency (the equiva- species and a means of distinguishing committee). The revision would include a lent of a journal’s editorial board). among them; to this core has been added the traditional description of each taxon and the However it worked, standards would exercise of resolving their evolutionary rela- location of type material. It might also need to be set and monitored by the interna- tionships. I believe that taxonomy needs to include material not currently required in a tional committee, who would also deter- expand to include other aspects of the formal description, for example keys and, for mine which institute houses which taxono- species’ biology, to become an information many groups, photographs or other illustra- my, and would prevent duplication of effort. science that curates our accumulated knowl- tions. For some organisms a gene sequence edge of that species in the way a gene annota- might be required. It would also include a Advantages tion in a genome database organizes our treatment of existing known synonyms to I believe that what I have described is evolu- knowledge of a particular protein. Third, I preserve contact with the older literature. tionary rather than revolutionary in that it think it is essential that the unitary taxono- This draft first web revision would be placed preserves the hard-won successes of current my of different groups evolves from the on the web for comments from the commu- taxonomy while dispensing with the histori- present taxonomy. We must preserve the nity, then after changes have been made in cal baggage. It is also evolutionary in that achievements of 250 years of distributed response, it would become the unitary tax- groups would move to the new unitary taxon- taxonomy, dispensing with the bad legacy of onomy of the group. omy as resources became available. It would the past but retaining the good. What would this mean? First, from this set a series of achievable targets that could be To illustrate how this could be done I time onwards all future work on the group used to spur major funding initiatives, for need refer only to the set of species in the first example the first web revision of mosquitoes, web revision and then later to those in the reptiles or plants (and I hope Nature or Box 1: Taxonomy ‘nth (that is, current) web revision’. The tax- Science might celebrate these milestones as on the web onomy of the group is thus at a stroke liber- they do completed genome sequences). ated from nineteenth-century descriptions I believe that major government and The current codes of zoological and botanical and potentially undiscovered synonyms. If I private research funders would consider nomenclature do not allow original descriptions to think I have discovered a new species I need construction and maintenance of a unitary be made purely on the web, but nevertheless only to check that it is not already in the web taxonomy — universally accessible, and the there is a substantial amount of taxonomy on the revision. So what happens if I describe a new foundation of all future work on the group — Internet. The Natural History Portal of the Natural species and then someone discovers that much more attractive to support than taxon- History Museum in London (www.nhm.ac.uk/ Linnaeus or someone had already described omy as presently practised. It might also portal/index.html) provides an excellent entry into it in an overlooked work? Well, that interest- attract new sources of funding. It surely isn’t these resources, which include such sites as the ing nugget of historical information can be impossible that a major company might International Plant Name Index (www.ipni.org) added to the species’ web page, but the name sponsor the web revision of, say, the Lepi- that covers all higher plants; the ant database doesn’t change. What happens if I want to doptera (butterflies and moths); and if it (www.antbase.org) featured recently in Nature’s lump, split or add species, or revise their wants to put its logo on the site, then why not? News section (416, 115; 2002); and the Tree of higher classification? Then I submit a revi- The web revision would become an infor- Life project (tolweb.org/tree), a database sion that is mounted on the web for referee- mation hub, both through its contents and of phylogenies. ing and comment. If, as a result, it is accept- through its links to other sites. Links to mole- The most common data available are catalogues ed, it becomes incorporated into the current cular databases will facilitate the increasing of species names and lists of museum specimens, (n+1th) web revision. At any one time there usefulness of molecular techniques in although some identification keys and other is just a single current web revision to which species identification. There are already information-rich sites are becoming available. people refer, linked to all previous revisions exciting web-based phylogenetic projects An ambitious project led by Species 2000 (which are maintained on the web, so that in (see Box 1) that aim ultimately to build a (www.sp2000.org) and the Integrated Taxonomic future I can easily see what was understood phylogeny of all living organisms; clearly, Information System (www.itis.usda.gov) aims to by species x in year y). one would build in reciprocal links to these catalogue the world’s biota, and these sites A major difference between this way of sites. Today, a reference to a species in a scien- themselves also link to the Global Biodiversity doing taxonomy and the status quo is that a tific article usually gives just the scientific Information Facility (www.gbif.org), intended to be a unitary taxonomy needs administration: name and possibly the authority, but seldom general clearing house for biodiversity information. both the physical implementation on servers refers (or gives credit) to the taxonomic revi- Finally, the All Species Foundation and networks, and the intellectual adminis- sion upon which the identification is based. (www.all-species.org) has set itself the goal of tration of the current web revision. One virtue As increasing numbers of journals go elec- making an inventory of all species on Earth in of the present system is that if no one is inter- tronic, the mention of a species can more and the next 25 years. ested in a group’s taxonomy it can quietly more easily be linked to its position in the slumber in the library. But the collections and current web revision. Were the status of the

species to change, the link would take you to and thus attracting people and funds into the the contemporary web revision and then for- field. But is such a root-and-branch change ward to the current conception of the taxon. in the culture of taxonomy really needed? These links could also be used to produce a Although there is near-universal agreement much-needed, fair ‘citation count’ for taxon- about the current depressed state of descrip- omists. Finally, as an increasing amount of tive taxonomy, wouldn’t more funding alone the scientific literature becomes available solve the problem? online through projects such as JSTOR I think not: indeed, descriptive taxonomy (www.jstor.org), one can imagine links might disappear completely for ‘difficult’ NATURAL HISTORY MUSEUM, LONDON HISTORY NATURAL between a species description and important groups such as many insects and nematodes. early papers on its taxonomy and biology, Just as Moore’s law says that microprocessor again maintaining links with the good legacy power doubles every 18 months, there must of distributed taxonomy. be a parallel law that says DNA sequencing Many taxonomic works are very hard for Harnessing the power of the web would allow all power increases geometrically. In 10 or non-specialists to use, sometimes because of contributions to taxonomy to be collated. 20 years’ time it will be simpler to take an real difficulties in telling many species apart, individual organism and get enough but more often because of the telegraphic my should also be stored on the web. Even if sequence data to assign it to a ‘sequence clus- jargon and lack of illustration imposed on they are not incorporated in the current web ter’ (equivalent to species) than to key it taxonomists by the expense of publication in revision they can at least influence future down using traditional methods, let alone print. The web has far fewer constraints, and scholarship and research. describe it as new. Just as bacterial taxonomy provides the space needed for taxonomists to An important issue is the degree to which is now nearly all sequence-based, a new way be understood. Taxonomy often pays insuffi- a treatment should be ‘complete’ before it is a of classifying insects, nematodes and per- cient attention to its ‘end users’, the ecologists, candidate for a first web revision. Could a haps even many plants and fish might evolve conservationists, pest managers and amateur series of intractable species complexes that is totally divorced from current taxono- naturalists who need or want to identify ani- requiring detailed research delay completion my — a point also made forcibly by Robert mals and plants. I hope that, overlaid on the of a revision? The ideal solution would be to May, president of Britain’s Royal Society. current web revision, there would be higher- commission new taxonomic research to sort Would the death of large swathes of pre- level information, the equivalent of the out these problems, but if this is not possible sent-day systematics matter? Yes it would, regional field guides and floras used by field I would favour a category of ‘provisional because we would be throwing away so much workers. For many, this ‘entry level’ would be taxon’, where the need for further study is of what we have learned in the past 250 years all that is required, but where needed the user clearly highlighted. After all, the hetero- about the planet’s biota, a lot of which we could burrow deeper, right through to the chromatin-rich gaps in the human genome would then have to relearn. But unless taxon- primary taxonomic sources. Today, few peo- sequence did not delay the announcement of omy is unitary, web-based and able to ple would seriously think about taking a com- its ‘completion’. accommodate these radical new ways of puter into the field as a substitute for a field Is a web-based taxonomy as permanent as doing biology, I fear it will be sidelined. guide, but that will undoubtedly change and a paper-based one, and are people without The rigidity built into the current rules taxonomists should be ready. computers disenfranchised, especially those and codes of taxonomy — which include pro- Finally, the taxonomy should be available in less wealthy countries? I believe the first is hibition of purely electronic description — is free (without access charges) to anyone who a non-issue; there is not (as far as I know) a part of their success, and changes should not can log onto the Internet. This will raise the paper back-up to the human genome data- be made lightly. But I suspect these rules are profile of taxonomy and increase the number base, and the international committee would now a brake on progress, imprisoning the of people who actually use the fruits of taxo- set rigid standards for archiving and backup. subject in outdated methodologies, and ren- nomic research. Longer-term positive bene- Access is a much more important matter, but dering it difficult or impossible to attract the fits will be for a new, young generation of very many more people are at present dis- major funds needed to reverse its slow naturalists, stalking their prey using digital enfranchised by their inability to get to a spe- decline. Surely it is time to experiment — cameras, downloading their captures into cialist library, or to order a reprint, or even by time for the international taxonomic com- PCs, then identifying them over the web — being unaware that certain literature exists. munity to come together and countenance a exposing them to taxonomy as an active dis- The web-based taxonomy must be com- unitary web revision of one or a few major cipline, at the heart of modern biology. pletely downloadable so that even continu- groups of organisms (and to work out exactly ous access to the Internet is not essential, how a unitary taxonomy should operate). Disadvantages and, if all else fails, a paper copy could be This venture must be sanctioned and sup- One disadvantage of a unitary taxonomy is printed. It might spread the geographical ported by the existing international commit- the requirement for more administration, distribution of taxonomic activity if some tees, or no serious taxonomist will waste his with its attendant costs. My assertion is that sites were hosted by developing countries or her time on it; no institution will adminis- the advantages of a unitary taxonomy will with strengths in computing, such as India. ter it; and no agency will fund it. If successful, prime sufficient new funds to counter- it will change how taxonomy is done for ever; balance this, but if I’m wrong the project Conclusions if it fails it would not be difficult to revert to fails. There are also considerable technologi- I find that the commonest reaction of taxon- the status quo ante. There is everything to gain cal challenges in developing the web software omists to these ideas is the worry that it is an and little to lose. ■ to support the taxonomies. attempted technological fix that distracts H. Charles J. Godfray is at the NERC Centre for A possible criticism is that the proposal is attention from what they (and I) perceive to Population Biology, Department of Biological top-down, at variance with the individualis- be the overwhelmingly critical issue — the Sciences, Imperial College at Silwood Park, Ascot, tic tradition of taxonomy. Would one clique lack of people and resources devoted to Berkshire SL5 7PY, UK. be able to impose its view of how a group is descriptive taxonomy. The counter-argu- classified? The international committee ment is that the technological fix is not an Acknowledgements would be empowered to set standards, but end in itself; it is the means of making grass- I am grateful to the many taxonomists and other biologists rejected contributions to a group’s taxono- roots taxonomy more accessible and useful, who have debated these issues with me.