Sir

The overwhelming task for taxonomy in ecological and biodiversity research arguably requires entirely new approaches. Web-based technology would be a great step towards a more accessible and universal platform, as proposed by H. C. J. Godfray in his Commentary1, and backed by F. A. Bisby in Correspondence2. But some crucial problems remain, not least the quality and accuracy of submitted information3. Many of these problems could be solved if DNA sequences were used as the universal reference standard.

Their usefulness for taxonomic purposes is not disputed. However, all current approaches use DNA as just an additional criterion for identifying a species or a taxon, without attempting to give it a central role. We believe that DNA taxonomy can provide a new scaffold for our accumulated taxonomic knowledge and a reliable tool for species identification and description.

DNA sequences alone are not sufficient to characterize a species4, but their unique reproducibility helps to guard against duplicate descriptions3. Moreover, collection and curation of extracted DNA samples is technically easy. DNA is very stable and any sample can be split into multiple subsamples, which can be sent to other museums as backups. DNA collections are already needed for the many projects in different laboratories looking at the phylogeny or phylogeography of species. These projects often involve very valuable samples, yet there is no scheme to safeguard them for future reference.

DNA sequencing is often considered to be a complex, expensive technology. But training good taxonomists is also very expensive, and it is a waste of resources to use them only for routine identification of specimens collected in research projects. We need good taxonomists to work on the huge task of matching existing taxonomic information with DNA sequences of new specimens, and to recognize and describe new species. Routine species identification should be the task of specialized DNA sequencing facilities, for which machines are readily available. A DNA facility that could routinely handle about 1,000 samples a day would cost approximately as much as a facility running a transmission and a scanning electron microscope.

We believe that a system based on DNA taxonomy can now be built to integrate the strengths of the traditional system with new technological possibilities, making full use of the invaluable information accumulated over the centuries and giving well-trained taxonomy specialists the opportunity to convert their expertise into broadly reproducible knowledge. We are developing these ideas into a longer article to be published elsewhere soon.