Preprints of the Metadiversity Conference Proceedings

  Session 2: The Challenge in Species Discovery and Taxonomic Information

Conventions, Standards, and Consensus in Systematic Practice: How Far Can (or Should) We Go?

PETER F. STEVENS, Professor of Biology, University of Missouri, St. Louis, and Curator, Missouri Botanical Garden

"The nice thing about standards is that you have so many to chose from; furthermore, if you do not like any of them, you can just wait for next year’s model" (Tanenbaum 1981, 168).

The goal of this talk is to begin to explore the general issue of the role of standards, conventions and metadata in systematic practice in general and species and their description and delimitation in particular. After a few necessary definitions, I will talk very briefly about the activities of the Taxonomic Databases Working Group (TDWG) in promoting standards and developing metadata. I then will suggest that thinking about conventions and standards helps illuminate the nature of plant taxa, and in particular the constraints on how species are delimited. I then will show how TDWG standards and convention are integral to the developing International Plant Names Index (IPNI) a database of plant names (initially flowering plants only). IPNI, with its bibliographic information on plant names, kept current and enabling linkages to other kinds of data, provides metadata useful in any biological project that uses the names of flowering plants. The standards developed, and others, can be used to simplify many aspects of systematic practice. Although we should try to standardize much of the routine of systematic work, the use of inappropriate standards can make answering some kinds of systematic questions difficult.

Standards, Conventions, Consensus and Metadata

First, some necessary definitions. A standard, to quote the Oxford English Dictionary is an "examplar of measure or weight"–hence "an authoritative or recognized exemplar of correctness, perfection, or some definite degree of any quantity." A convention, on the other hand, is "a rule or practice based on general consent." The word convention quite often has negative connotations–it is artificial, arbitrary and formal, while a standard refers to something in the real world, or at least something we think is more important. Consensus, "the collective unanimous opinion of a number of persons," is one way in which standards and conventions become effective. Standards in particular come in a variety of flavors, and we can usefully distinguish here between de facto, de jure, and formal standards (the last are standards accepted by a standards body like the International Standards Organisation (cf. Mowbray & Malveau 1997)). Standards that deal with "correctness" can be called reference standards; if they refer to rules or actions necessitated by a body of systematic theory, they can be called rule-based standards; and if they deal with "some definite degree of any quantity," they can be called metric standards. These distinctions are not exclusive, but they do emphasize how heterogeneous "standards" are. Metadata are data about data, or data that underpin the use of data in a particular way; and standards and conventions are involved in the development of metadata.

The Taxonomic Databases Working Group (TDWG)

The International Working Group on Taxonomic Databases was set up in 1985 to establish international collaboration among database projects. Its goal is to aid in the dissemination and exchange of information by focusing on the common use and interpretation of terminology, data fields, dictionaries, common logical rules, and data relationships. Basically, it deals with metadata and standards. Initially botanical in focus, it has broadened out over the years, although there is a pressing need to become yet more active.

The role of TDWG is that of a facilitator, and it may evaluate and endorse existing standards or get groups of people together to make new ones. The kind of projects in which it has been involved include defining standards for data collection in economic botany (Cook 1995), geographic names (Hollis & Brummitt 1992), abbreviations for authors (Brummitt & Powell 1992), while it endorses such important standards as a format for recording and exchanging descriptive taxonomic data, DELTA (Descriptive Language for Taxonomy) (Dallwitz & Paine 1986), and the herbaria acronyms used when specimens are cited (Holmgren et al. 1990). Many such standards are in common use, greatly facilitating communication among biologists.

Of course, new authors describe new species for the first time, new herbaria are formed, and so new editions of these standards are produced, or, for the Web versions, they are kept current–and for our future efforts, that is where the emphasis will be. Indeed, with the decentralization now possible with such projects (see below), we need not get into the situation of having an out-of-date standard or a standard that is independently improved by different groups–parallel evolution here, as elsewhere, is the bane of systematics! As the need arises, we can develop new–and extend the coverage of–existing standards. Is there confusion in how botanists and zoologists cite collections? What about author and journal abbreviations and contractions across all groups, not just plants? As lists of taxa (this refers to groups at any level of the taxonomic hierarchy, from kingdom to variety) become ever more comprehensive, these are issues that we may need to consider. Codes of nomenclature are also standards and metadata, and again, there is a need to establish whether there is any call for a unified code of biological nomenclature.

Conventions, Standards, and Systematics

I begin by thinking of the higher levels of the taxonomic hierarchy–classes, families, genera, etc.–in the context of standards and conventions. If taxa at the same rank, e.g., a genus, were equivalent, this might provide some kind of metric for measuring diversity. However, for most systematists higher taxa are simply groups that contain one or usually more lower taxa. They may delimit groups that a systematist for some reason or other thinks are particularly interesting, but that does not convert to any equivalence of groups at the same hierarchical level (Stevens 1997a). There is widespread acceptance of the principle of monophyly, and this means that all higher taxa have to be monophyletic, i.e., contain all and only the descendents of a particular common ancestor, but this is a grouping, not a ranking criterion. However, not only is there still no consensus here, but monophyly is anyway a rule-based, not a metric, standard. Given the constraints of monophyly, it is convention driven by a need to communicate that leads us to give a named group a particular circumscription; we apply names at ranks like order, family, and genus to those parts of a taxonomy that are commonly used in teaching and in general communication between biologists (Angiosperm Phylogeny Group 1998: a phylogenetic naming system will not escape this need to act as a communication system), and so they are in the context a kind of reference standard. But whatever one’s systematic philosophy, there is no measurement unit inherent in a biological classification. If I walk two miles, I have unarguably walked twice as far as if I had walked one mile; if my state has 500 genera of plants and yours has 1000, I cannot say that your state is twice as diverse as mine. Similarly, although it may be something of a convention to use numbers of species in estimates of diversity and amount of radiation (systematists have suggested ways of quantifying these), they, too, are a poor metric for this.

What about conventions and these species? Here I touch both on the act of writing descriptions and on how we go about circumscribing species–and I should say that nearly all the descriptive systematic work I do is on tropical plants.

Writing Descriptions

Linnaeus used a brief, almost telegraphic style for his descriptions, and he defined and illustrated the terms he used (Linnaeus 1751). Over the years Linnaeus’s definitions of shape terms and the like were modified and alternative usages crept into use. Definitions of some shape terms were reviewed by a committee formed under the aegis of the Systematics Association and their consensus published (Anonymous 1962) and adopted by Stearn (1992) in his invaluable "Botanical Latin." These have become reference standards, standards for communication, to be used as a matter of good practice by all systematists when describing plants. Unfortunately, they neither map onto discontinuities in nature nor provide a metric for shape. They are imprecise and even positively misleading if it is assumed that two taxa described as having, say, a differently-shaped lamina apex are thus necessarily assignable to two discretely different shapes, or if it assumed that two taxa with the same shape necessarily do not have discretely different shapes.

That the definitions of these shape terms, although accessible, are difficult to apply without reference to diagrams–yet they generally are so used–points to a problem with many reference standards. Accessibility and ease of use are essential for their adoption, for them to become standards de facto. This is particularly true of color standards, such as the set devised by Ridgway (1913) and published in a single edition of 4,000 copies. Its problems were evident by the middle of the century (Hanly 1949), but it is still in use, partly because Ridgway focused on colors common in the living world. But other color standards have inevitably (given the printing history of Ridgway’s book, and the fact that the colors there are not all stable) come into use, and these include the Munsell and Royal Horticultural Society color charts, that proposed for British fungi (Henderson et al. 1969), and so on. These later standards are more accessible, although they may focus on different parts of the spectrum. In any case, many biologists prefer to use terms like "red," "violet," and "lilac"–terms in common use but without precise definitions.

Units for measuring size may be arbitrary in the sense that there are all sorts of units used by different cultures or groups. However, inches and centimeters are interconvertible, and they provide a metric by which measurements can be related to one another and manipulated in a way that is impossible with many biological descriptors, which are ultimately reference standards. Biological descriptors refer to a variety of kinds of objects, the extremes of which are terms like "ovule" and "radius"–structures that are thought to have arisen only once in the course of evolution–and "wing", "stipule", and "campylotropy"–for which there is no such evidence. The former often have more precise definitions and are commoner in vertebrate zoology than in flowering plants (e.g., cf. Peters 1964; Wheeler et al. 1993). Reference terms like "acute" are widely used, although not necessarily with identical definitions.

Circumscribing Species

I find it unsettling that many of the operations we carry out as we look at specimens and estimate variation patterns–that is, much of systematic practice–amount to a convention, and a largely unanalyzed one at that (cf. Stuessy 1994, who stresses the development of "skills of intuitive [sic] pattern recognition," albeit in the context of well-studied temperate groups). The operational species concept many of us use (by "us," I mean systematists in general) is "a species is what a competent systematist says it is" (Regan 1926). This concept–or perhaps more accurately, convention–has been with us for over 150 years (e.g., Stevens 1991, 1997b). In practice this concept/convention has often been used to promote broadly delimited species, with a reduction in the numbers of species recognized being one of its useful results–a rather strange argument, one would have thought. Species are also explicitly based on discontinuities in variation pattern, and there are conventions for how much and what kind of variation is needed to indicate different levels in the taxonomic hierarchy at and below the level of species (e.g., Davis & Heywood 1963; Mayr & Ashlock 1981). But there is disagreement over such conventions, and subspecies and varieties may be differently defined by systematists in different countries (Hamilton & Reichard 1992), or the significance of variation ornithologists had previously called subspecific may be questioned (see Cracraft 1992, on the numbers of species of birds of paradise). It is still more unsettling to one who has long wished for closer links between academia and the museum to observe the lack of connection between those who theorize about the nature of species, a very active area and largely in academia, and those who describe species (McDade 1995), an area that is somewhat in regress and is largely carried out in museums.

A monographer is privileged–he or she can analyze material from throughout the range of a group, judiciously evaluate the variation, and make decisions about the limits of taxa with the "big picture" in mind. But without clarification of criteria for collecting and evaluating data on variation, this privilege amounts to little. I have recently been looking at a widely ranging species of Rinorea (Violaceae) growing from Myanmar and Hainan to Borneo (Jarvie & Stevens 1998; cf. Jacobs & Moore 1961). There is not one, but three species–one with two subspecies, and another two species are still to be described. Similarly, in a recent treatment of Fagraea (Gentianaceae), 42 species from Borneo are recognized, of which 24 are endemic, while in an earlier treatment the corresponding figures are 14 and 3 (Wong & Sugau 1996; cf. Leenhouts 1962). Importantly, in both cases material seen by the earlier authors was assigned by the later authors to different, and in Fagraea, often newly described species; little that was completely new had turned up in the intervening years. But if I ask the questions, Who is right? and why are there differences? it is remarkably difficult to provide answers. We are not sure of whether or not we are delimiting species for a herbarium taxonomist, and we are not sure of the significance of finding locally discontinuous variation that appears to break down in a geographically broader study (e.g., Stevens 1993; Cronk 1999). All too often we seem not to know why (for whom?) and how we should analyze variation (critical comments made by Heywood 1974, about taxonomy are still, unfortunately, all too relevant).

The answers to such questions go beyond the kind of convention and standard I have been discussing, although they start there. We desperately need to articulate what is involved in taxonomic practice if the cherished independence of systematists–the wish each of us has for our own opinions on the limits of species to be taken seriously–is to amount to much. We must work towards clarification of how we collect and analyze data, articulating reference and rule-based standards where necessary, and clarifying and justifying the role convention plays in the whole systematic process.

The International Plant Names Index

Although at the level of species delimitation we are dealing both with unruly nature and unruly naturalists, there are less problematical areas where standards, conventions, and metadata can facilitate the speed and ease with which we do our basic systematic work. The Plant Names Project and its main product, the International Plant Names Index (IPNI), serves to illustrate this.

The names in IPNI will be those currently in the Index Kewensis, based at the Royal Botanic Gardens, Kew; the Gray Herbarium Index, at Harvard; and the Australian Plant Names Index, at Canberra. IPNI will initially include well over 1,000,000 names of flowering plants, and we hope to extend the coverage to all vascular plants in the current phase of the project. In short, it aims to be an authoritative list of all plant names and their place of publication, kept current, freely available and query-able, decentralized, and maintained with a minimum of bureaucracy. It will of course take some time to meet all these ambitious goals. (Further details of the project may be found in Croft et al. (submitted) http://pnp.huh.harvard.edu.)

The names will reside in a distributed database, and although only the three institutions just mentioned are involved at present, IPNI is designed to be highly scalable. All new names appearing (including citations of types specimens–types are themselves standards that allow us to relate names to plants with the minimum of ambiguity) will be added from Kew, Harvard, or Canberra and will be accessible as soon as they are added. TDWG standards for geography, abbreviations, and contractions of names and journals, etc., will be followed; these standards, too, will be kept current and generally accessible.

Of considerable importance is the submissions module that has been developed. There are four aspects of the submissions mechanism that I want to mention here. First, although these changes will appear in IPNI as soon as they are made, it will be clear to all users that they have not been checked. Editors, whether of particular taxa, particular types of names (e.g., names described by Linnaeus), or of particular directories (e.g., periodical lists), will check the changes. Second, any and all changes suggested will become part of the permanent contribution history of that record, and this history, too, can be accessed by users. Third, by making it easy for the whole community to submit additions and corrections, the difficult process of editing and verifying the existing entries will be greatly facilitated. Fourth, the submission module itself can be adapted for a variety of uses, so it, too, can become a kind of standard. For instance, it could serve as the mechanism by which names proposed for a phylogenetic naming system are recorded.

IPNI is a nomenclatural list, including the original citation of a name and, where possible, that of its type specimen. Any name referring to vascular plants should be traceable in it, and for most purposes the information it contains will function as metadata, serving to ground names in the literature. To IPNI can be linked images of the protolog, including any illustrations, images of the type, etc. It can be mirrored to institutions and projects, and all projects linked with it will be synchronized with it and kept up-to-date with the information it contains, while corrections to it made in the course of work on any of these projects can be made immediately accessible to all.

Conclusions

Much of the work of describing species and writing floras and checklists is almost formulaic. Reference standards and conventions for journal abbreviations, citation of type and other specimens and of synonymy, etc., are very desirable, and their use should be promoted by groups like TDWG, professional societies, etc. They certainly should become de facto standards, if not formal standards, and failure to use them leads to the waste of an inordinate amount of time by systematics. Reference standards may be arbitrary. There may be more than one way of handling journal and book abbreviations and contractions, but so long as the way adopted works, its principles are explicit, and the consequences of its adoption are understood. That is all that matters. Until fairly recently it was a rite of passage for all new flora projects to argue interminably about how names should be handled–should the Australian botanist Ferdinand von Mueller be F. von Muell., F. Mueller, F. v. M., or something else? With a list of author names and abbreviations, this need no longer happen; we can use either a full name or a single contraction. The fewer standards covering such aspects of a species description, the more our work will be speeded up; this is not an area where having several standards is a good thing (cf. the epigraph). Simplification and standardization not only speeds up our work, but it makes it much easier for a student to understand what we are doing. But there are circumstances in which the use of a particular standard may be inappropriate. For writing a flora, the use of plain language and of reference standards such as those in Stearn (1992) is essential, but metric standards are needed for actually deciding on the limits and relationships of species. Here more than one standard is involved. (Note that metric standards can be converted accurately to reference standards, but the reverse operation may well introduce imprecision (see also Heywood 1984).)

I have barely begun to explore the role and interaction of standards, conventions, and metadata in systematics. However, thinking of such issues in the context of systematic practice will help us to understand better what we do, to gather appropriate data more quickly, and will allow meaningful consensus over the limits of the taxa we use to describe the living world–and this makes the important link to projects like ITIS, BONAP, and Species 2000.

Acknowledgements

I thank W. L. Alverson, D. E. Boufford, J. Cadle, N. Cross and F. Pando de la Hoz for helpful discussion. The support of the NSF and USGS (NSF DEB-9726045 and DBI-9808220) is gratefully acknowledged.

References

Angiosperm Phylogeny Group. 1998. An ordinal classification for the families of flowering plants. Ann. Missouri Bot. Gard. 85:531-553.

Anonymous. 1962. Systematics Association committee for descriptive biological terminology II. Terminology of simple symmetrical plane shapes (Chart 1). Taxon 9:145-156, 245-247.

Brummitt, R. K., and C. E. Powell. 1992. Authors of plant names. Royal Botanic Gardens, Kew.

Cook, F. E. M. 1995. Economic botany data collection standards. Royal Botanic Gardens, Kew.

Cracraft, J. 1992. The species of the birds-of-paradise (Paradisaeidae): Applying the phylogenetic species concept to a complex pattern of differentiation. Cladistics 8:143.

Croft, J., N. Cross, E. Nic Lughadha, P. F. Stevens, J. G. West, and G. Whitbread. Plant names for the 21st century: The International Plant Names Index, a distributeddata source of general accessibility. Taxon, accepted.

Cronk, Q. C. B. 1998. The ochlospecies concept. Ch. 11 in: Huxley, C.R., J. M. Lock, and D. F. Cutler (eds), Chorology, Taxonomy and Ecology of the African and Madagascan Floras. Kew Bulletin Additional Series, H.M.S.O., London.

Dallwitz, M. J., and T. A. Paine. 1986. User’s guide to the DELTA system. Ed. 4. Division of Entomology, Commonwealth Scientific and Industrial Research Organisation, Australia.

Davis, P. H., and V. A. Heywood. 1963. Principles of angiosperm taxonomy. Oliver & Boyd, Edinburgh.

Hamilton, C. W., and S. H. Reichard. 1992. Current practice in the use of subspecies, variety and form in the classification of wild plants. Taxon 41:485-498.

Hamly, D. H. 1949. Robert Ridgway’s color standards. Science 109:605-608.

Henderson, D. M., P. O. Orton, and R. Watling. 1969. British fungus flora. Vol. 1. H.M.S.O., Edinburgh.

Heywood, V. H. 1974. Systematics–the stone of Sisyphus. Bot. J. Linn. Soc. 6:169-178.

Heywood, V. H. 1984. Electronic data processing in taxonomy and systematics. pp.1-15 in Allkin, R. and F. A. Bisby (eds), Databases in systematics. Academic Press, London.

Hollis, S. and R. K. Brummitt. 1992. World geographic scheme for recording plant distributions. Hunt Institute for Botanical Documentation, Carnegie Mellon University.

Holmgren, P. K., N. H. Holmgren, and L. C. Barnett. 1990. Index herbariorum. Part 1. The herbaria of the world. Ed. 8. New York Botanical Gardens, New York. Jacobs, M., and D. M. Moore. 1971. Violaceae. pp. 179-212 in van Steenis, C. G. G. J. (ed.), Flora malesiana. Series 1, Vol. 7. Wolters-Noordhoff, Groningen.

Jarvie, J. K., and P. F. Stevens. 1998. New species and notes on Violaceae and Flacourtiaceae from Indo-Malasia. Harvard Pap. Bot. 3:255-264.

Leenhouts, P. W. 1962. Loganiaceae. pp. 293-387 in van Steenis, C. G. G. J. (ed.), Flora malesiana. Series 1, Vol. 6. Wolters-Noordhoff, Groningen. Linnaeus, C. 1751. Philosophia botanica... Kiesewetter, Stockholm.

Mayr, E., and P. D. Ashlock. 1981. Principles of systematic zoology. Ed. 2.

McGraw Hill, New York.

McDade, L. A. 1995. Species concepts and problems in practice: Insights from botanical monographs. Syst. Bot. 20:606-622.

Mowbray, T. J., and R. C. Malveau. 1997. Corba design standards. John Wiley, New York.

Peters, J. A. 1964. Dictionary of herpetology. Hafner, New York.

Regan, C. T. 1926. Organic evolution. Rep. British Assoc. Adv. Sci. 1925, 75-86.

Ridgway, R. 1912 (1913). Color standards and color nomenclature. The Author, Washington, D.C.

Stearn, W. T. 1992. Botanical Latin. Ed. 4. David and Charles, Newton Abbott.

Stevens, P. F. 1990. Species: Historical perspectives. pp. 302-311 in Keller, E. F., and E. Lloyd (eds), Keywords in Evolutionary Biology. Harvard University Press, Cambridge, Mass.

Stevens, P. F. 1997a. How to interpret botanical classifications–suggestions from history. BioScience 47:243-250.

Stevens, P. F. 1997b. J. D. Hooker, George Bentham, Asa Gray and Ferdinand Mueller on species limits in theory and practice: A mid-nineteenth-century debate and its repercussions. Hist. Records Austral. Sci. 11:345-370.

Stuessy, T. F. 1994. Case studies in plant taxonomy: Exercises in applied pattern recognition. Columbia University Press, New York.

Tanenbaum, A. S. 1981. Computer networks. Prentice Hall, Englewood Cliffs, New Jersey.

Wong, K. M., and J. B. Sugau. 1996. A revision of Fagraea (Loganiaceae) in Borneo, with notes on related Malesian species and 21 new species. Sandakania 8:1-93.

Previous | Next

Copyright © 2003 NFAIS. All rights reserved. No part of this product or service may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written consent.

Privacy Policy