Sir

The proposal by Andrew Polaszek and colleagues in their Commentary article (Nature 437, 477; 2005) for a universal register for animal names as a requirement of the next edition of the International Code of Zoological Nomenclature is very welcome. The authors note that a system was tried out by botanists and approved in principle by the International Botanical Congress in Tokyo in 1993, but was rejected by the subsequent congress in St Louis, Missouri, in 1999 and never implemented.

However, a formal system of registration of names was first proposed by mycologists 50 years ago (G. C. Ainsworth & R. Ciferri, Taxon 4, 3–6; 1955). Many mycologists were disappointed at the St Louis rejection, and in response, they established MycoBank in 2004 (P. D. Crous et al. Mycol. Res. 108, 1236–1238; 2004). Authors can obtain a unique accession number for their new entities through the web from MycoBank.org before publication, much in the way GenBank accession numbers are requested. New descriptions with MycoBank numbers are already appearing in the fungal literature.

The system is voluntary at present, but it is anticipated that leading journals, now starting to recommend this practice, will make it a requirement for the acceptance of papers with new scientific names in them as it becomes better known — just as they already require GenBank numbers for molecular sequences.

MycoBank, led by the Centraalbureau voor Schimmelcultures fungal biodiversity centre in Utrecht, is being developed in close collaboration with Index Fungorum, an online database of 380,000 scientific names. MycoBank accession numbers are also being incorporated into the life science identifiers scheme.

MycoBank is sure to evolve and become more honed as use of it increases. It provides a model that zoologists and others interested in bringing increased order into the naming of life on Earth might examine.