Taxonomy users
Sarah Kemmitt
Thursday, 12 February 2009 10:04 UTC
I think a useful distinction to make is between taxonomists and users of taxonomies. Many ecologists, molecular biologists, parasitologists, pathologists and other scientists make use of taxonomies & classifications in their work, but are not experts in taxonomy itself. How could taxonomic data be mobilized more effectively for your community? What are the resources that you currently use and do they meet your needs?
-
Replies
-
I’d like to see a user requirements analysis for a particular aspect of taxonomy. What do taxonomists actually need to produce to meet the needs of, say, an ecologist carrying out meta analysis of a series of papers? Without quite detailed user requirements gathering it is difficult to say whether taxonomy is meeting needs or not. My hunch is that much of taxonomy/systematics is carried without the needs of the users you mention in mind.
-
One of the obligations of countries party to the Convention on Biodiversity is to carry out a ‘Taxonomic Needs Assessment’. This arises from the acceptance of the Global Taxonomy Initiative as a part of the CBD, its aim being to ensure that there is sufficient taxonomic expertise and information to enable implementation of the Convention. The UK carried out such an assessment in the context of conservation, which you can find here There are a few other country reports there also. We also carried out one globally in the context of Invasive Alien species Further information on how these have been carried out is at the BioNET site and the GTI All of these are very much what is needed by users of taxonomy and the implications of those needs rather than the needs of taxonomists per se
-
It is worth remembering that taxonomists have little or no incentive to meet the needs of users other than traditional taxonomy. This is simply an issue of how people are judged and get promotions/tenure/recognition in an academic environment. Until this changes, most taxonomists, no matter how good their intentions (and many taxonomists do try to meet the needs of other users), do not have time to put their results into a form that fills the needs of others outside of taxonomy.
Access to taxonomic information in ways that meet users’ needs is, in my opinion, the real issue. GBIF, TDWG and others are an excellent step forward to making specimen and name data more accessible, but there is a very long way to go.
Another example, taxonomic literature is full of information that meets the needs of the users and the user needs analyses mentioned in previous posts. This information, including specimen, name, and a great deal of hidden biological information, can be extracted and repurposed to meet those needs, but currently that takes sn inordinate amount of time and effort, which is not realistic for most users.
What is the answer? I believe the answer can be met by INOTAXA, which is taking digitized taxonomic literature and extracting the data into a database in such a way that it can be repurposed for a wide variety of users in formats useful to them. While the project is still in prototype form, demonstrations and testing with users have shown us that there is a huge need and desire for this kind of access by taxonomists and ecologists (and others, though we have not yet demonstrated that).
-
For me the key thing in Anna’s comment is:
This information, including specimen, name, and a great deal of hidden biological information, can be extracted and repurposed to meet those needs, but currently that takes an inordinate amount of time and effort…
Doing this manually on a large scale is crazy. Taxonomy (and biodiversity informatics in general) has been pretty slow to exploit the kind of text mining tools that are widely used in bioinformatics. Recognising taxonomic names in text is the one bright spot. But there’s a wealth of data that could be easily harvested. For example, the titles of papers that describe new parasites often mention the host organism. With minimal effort we could build databases of host-parasite associations by mining those paper titles.
-
Certainly in microbiology we need more defined standards for taxonomy. What I mean is, we need a list of “If you want to propose a new organism, you must do the following” including things like “sequence the rrs gene, sequence the gyrB gene, test if the organism ferments glucose, test if the organism can oxidise methanol” etc. As it stands, you “just” have to produce a valid argument that the organism is “new”, but there are no standard lists of properties you must test. A good example is that people will usually test a range of carbon sources the organism uses and will list them in the paper – but everyone uses a different set of carbon sources…so then when you’re trying to compare two species and one grows on “methanol, methylamine, carbon disulfide and formate” but the other grows on “yeast extract, phenol, peptone and cellulose” it’s like trying to compare apples and oranges. I’d like some kind of standardisation of this.
-
The traditional need of the end user of taxonomic research is, at its most basic, to be enabled to correctly identify a specimen to species. This ability is flatly critical. Dichotomous keys are fine as far as they go but given an end user with an unknown specimen at hand it can still be quite a challenge to successfully complete the identification if that user does not have sufficient experience with related specimens. There is often a requirement for a certain level of expertise that is gained only through diligent and time-consuming effort.
Expert systems could be valuable in encoding that expertise. For instance, this can take the form of training artificial neural networks. In the past, I was involved in a proof-of-principle study (Do MT, Harp JM, Norris KC. Test of a pattern recognition system for identification of spiders Bulletin of Entomological Research 89 , 217-224 (1999) ) that has since been taken up at the American Museum of Natural History. If taxonomists can be supplied with software to allow training of expert systems, then that expertise can be exported and could even become a part of publication of the results of taxonomic research.
-
I would analyze Joel’s statement a bit further, and paraphrase to say that the traditional need of the end user of taxonomic research is, at its most basic, to be able to communicate about a species, or collection of species. It is of course crucial that specimens can be given a name, but the main purpose of giving the name is to allow information to be transferred from one person to another. Very importantly, we need to be able to talk about groups of organisms generally because they share some property or other, like wasps so that you know to treat their sting with vinegar. I wonder whether modern taxonomy is helping users to identify groups in a way that facilitates communication.
-
